Upload 10 files
Browse files- model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +2187 -3
- training_args.bin +1 -1
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 598635032
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8fe5d6868ff2c81227fd2c969c46af4fb1f58973ee4e0966c10979962f78982d
|
| 3 |
size 598635032
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1197359627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:47b3c36d9ca1786cd8d0139a8f5f16980ca80bc74296c03c792d092074a01113
|
| 3 |
size 1197359627
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:de790fba095fbd174e2cbec99a78f74e882cac29a5ff6c7320d627732b666d8d
|
| 3 |
size 14645
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:18105fb0a6af67c0e74a6e195673b14e0b259cf81cc485bb090b6d14e24299a1
|
| 3 |
size 1465
|
trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 1000,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -6957,6 +6957,2190 @@
|
|
| 6957 |
"eval_samples_per_second": 196.535,
|
| 6958 |
"eval_steps_per_second": 1.542,
|
| 6959 |
"step": 89000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6960 |
}
|
| 6961 |
],
|
| 6962 |
"logging_steps": 100,
|
|
@@ -6976,7 +9160,7 @@
|
|
| 6976 |
"attributes": {}
|
| 6977 |
}
|
| 6978 |
},
|
| 6979 |
-
"total_flos":
|
| 6980 |
"train_batch_size": 128,
|
| 6981 |
"trial_name": null,
|
| 6982 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.036280622240579596,
|
| 6 |
"eval_steps": 1000,
|
| 7 |
+
"global_step": 117000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 6957 |
"eval_samples_per_second": 196.535,
|
| 6958 |
"eval_steps_per_second": 1.542,
|
| 6959 |
"step": 89000
|
| 6960 |
+
},
|
| 6961 |
+
{
|
| 6962 |
+
"epoch": 0.00027908170954291995,
|
| 6963 |
+
"grad_norm": 1.6284213066101074,
|
| 6964 |
+
"learning_rate": 4.331808372709617e-05,
|
| 6965 |
+
"loss": 2.1737,
|
| 6966 |
+
"step": 89100
|
| 6967 |
+
},
|
| 6968 |
+
{
|
| 6969 |
+
"epoch": 0.0005581634190858399,
|
| 6970 |
+
"grad_norm": 1.5599644184112549,
|
| 6971 |
+
"learning_rate": 4.33030093580447e-05,
|
| 6972 |
+
"loss": 2.1666,
|
| 6973 |
+
"step": 89200
|
| 6974 |
+
},
|
| 6975 |
+
{
|
| 6976 |
+
"epoch": 0.0008372451286287599,
|
| 6977 |
+
"grad_norm": 1.5871340036392212,
|
| 6978 |
+
"learning_rate": 4.328792063355065e-05,
|
| 6979 |
+
"loss": 2.1656,
|
| 6980 |
+
"step": 89300
|
| 6981 |
+
},
|
| 6982 |
+
{
|
| 6983 |
+
"epoch": 0.0011163268381716798,
|
| 6984 |
+
"grad_norm": 1.655039668083191,
|
| 6985 |
+
"learning_rate": 4.327281756544842e-05,
|
| 6986 |
+
"loss": 2.169,
|
| 6987 |
+
"step": 89400
|
| 6988 |
+
},
|
| 6989 |
+
{
|
| 6990 |
+
"epoch": 0.0013954085477146,
|
| 6991 |
+
"grad_norm": 1.549782633781433,
|
| 6992 |
+
"learning_rate": 4.325770016558367e-05,
|
| 6993 |
+
"loss": 2.1748,
|
| 6994 |
+
"step": 89500
|
| 6995 |
+
},
|
| 6996 |
+
{
|
| 6997 |
+
"epoch": 0.0016744902572575198,
|
| 6998 |
+
"grad_norm": 1.6504789590835571,
|
| 6999 |
+
"learning_rate": 4.3242568445813306e-05,
|
| 7000 |
+
"loss": 2.1637,
|
| 7001 |
+
"step": 89600
|
| 7002 |
+
},
|
| 7003 |
+
{
|
| 7004 |
+
"epoch": 0.00195357196680044,
|
| 7005 |
+
"grad_norm": 1.6154310703277588,
|
| 7006 |
+
"learning_rate": 4.322742241800545e-05,
|
| 7007 |
+
"loss": 2.1582,
|
| 7008 |
+
"step": 89700
|
| 7009 |
+
},
|
| 7010 |
+
{
|
| 7011 |
+
"epoch": 0.0022326536763433596,
|
| 7012 |
+
"grad_norm": 1.611567497253418,
|
| 7013 |
+
"learning_rate": 4.321226209403947e-05,
|
| 7014 |
+
"loss": 2.1519,
|
| 7015 |
+
"step": 89800
|
| 7016 |
+
},
|
| 7017 |
+
{
|
| 7018 |
+
"epoch": 0.0025117353858862797,
|
| 7019 |
+
"grad_norm": 1.5689648389816284,
|
| 7020 |
+
"learning_rate": 4.319708748580592e-05,
|
| 7021 |
+
"loss": 2.1659,
|
| 7022 |
+
"step": 89900
|
| 7023 |
+
},
|
| 7024 |
+
{
|
| 7025 |
+
"epoch": 0.0027908170954292,
|
| 7026 |
+
"grad_norm": 1.630550503730774,
|
| 7027 |
+
"learning_rate": 4.318189860520658e-05,
|
| 7028 |
+
"loss": 2.1545,
|
| 7029 |
+
"step": 90000
|
| 7030 |
+
},
|
| 7031 |
+
{
|
| 7032 |
+
"epoch": 0.0027908170954292,
|
| 7033 |
+
"eval_loss": 2.2889482975006104,
|
| 7034 |
+
"eval_runtime": 51.8104,
|
| 7035 |
+
"eval_samples_per_second": 196.756,
|
| 7036 |
+
"eval_steps_per_second": 1.544,
|
| 7037 |
+
"step": 90000
|
| 7038 |
+
},
|
| 7039 |
+
{
|
| 7040 |
+
"epoch": 0.00306989880497212,
|
| 7041 |
+
"grad_norm": 1.5548030138015747,
|
| 7042 |
+
"learning_rate": 4.316669546415441e-05,
|
| 7043 |
+
"loss": 2.1585,
|
| 7044 |
+
"step": 90100
|
| 7045 |
+
},
|
| 7046 |
+
{
|
| 7047 |
+
"epoch": 0.0033489805145150396,
|
| 7048 |
+
"grad_norm": 1.675545573234558,
|
| 7049 |
+
"learning_rate": 4.315147807457356e-05,
|
| 7050 |
+
"loss": 2.1588,
|
| 7051 |
+
"step": 90200
|
| 7052 |
+
},
|
| 7053 |
+
{
|
| 7054 |
+
"epoch": 0.0036280622240579597,
|
| 7055 |
+
"grad_norm": 1.6402935981750488,
|
| 7056 |
+
"learning_rate": 4.313624644839936e-05,
|
| 7057 |
+
"loss": 2.1548,
|
| 7058 |
+
"step": 90300
|
| 7059 |
+
},
|
| 7060 |
+
{
|
| 7061 |
+
"epoch": 0.00390714393360088,
|
| 7062 |
+
"grad_norm": 1.6139987707138062,
|
| 7063 |
+
"learning_rate": 4.312100059757829e-05,
|
| 7064 |
+
"loss": 2.1571,
|
| 7065 |
+
"step": 90400
|
| 7066 |
+
},
|
| 7067 |
+
{
|
| 7068 |
+
"epoch": 0.0041862256431437995,
|
| 7069 |
+
"grad_norm": 1.625307321548462,
|
| 7070 |
+
"learning_rate": 4.310574053406801e-05,
|
| 7071 |
+
"loss": 2.1595,
|
| 7072 |
+
"step": 90500
|
| 7073 |
+
},
|
| 7074 |
+
{
|
| 7075 |
+
"epoch": 0.004465307352686719,
|
| 7076 |
+
"grad_norm": 1.6252094507217407,
|
| 7077 |
+
"learning_rate": 4.3090466269837304e-05,
|
| 7078 |
+
"loss": 2.1593,
|
| 7079 |
+
"step": 90600
|
| 7080 |
+
},
|
| 7081 |
+
{
|
| 7082 |
+
"epoch": 0.00474438906222964,
|
| 7083 |
+
"grad_norm": 1.6178643703460693,
|
| 7084 |
+
"learning_rate": 4.307517781686611e-05,
|
| 7085 |
+
"loss": 2.1393,
|
| 7086 |
+
"step": 90700
|
| 7087 |
+
},
|
| 7088 |
+
{
|
| 7089 |
+
"epoch": 0.005023470771772559,
|
| 7090 |
+
"grad_norm": 1.4477986097335815,
|
| 7091 |
+
"learning_rate": 4.3059875187145495e-05,
|
| 7092 |
+
"loss": 2.1526,
|
| 7093 |
+
"step": 90800
|
| 7094 |
+
},
|
| 7095 |
+
{
|
| 7096 |
+
"epoch": 0.00530255248131548,
|
| 7097 |
+
"grad_norm": 1.6491955518722534,
|
| 7098 |
+
"learning_rate": 4.3044558392677627e-05,
|
| 7099 |
+
"loss": 2.1456,
|
| 7100 |
+
"step": 90900
|
| 7101 |
+
},
|
| 7102 |
+
{
|
| 7103 |
+
"epoch": 0.0055816341908584,
|
| 7104 |
+
"grad_norm": 1.6281492710113525,
|
| 7105 |
+
"learning_rate": 4.30292274454758e-05,
|
| 7106 |
+
"loss": 2.1516,
|
| 7107 |
+
"step": 91000
|
| 7108 |
+
},
|
| 7109 |
+
{
|
| 7110 |
+
"epoch": 0.0055816341908584,
|
| 7111 |
+
"eval_loss": 2.270752191543579,
|
| 7112 |
+
"eval_runtime": 51.3993,
|
| 7113 |
+
"eval_samples_per_second": 198.329,
|
| 7114 |
+
"eval_steps_per_second": 1.556,
|
| 7115 |
+
"step": 91000
|
| 7116 |
+
},
|
| 7117 |
+
{
|
| 7118 |
+
"epoch": 0.005860715900401319,
|
| 7119 |
+
"grad_norm": 1.6257696151733398,
|
| 7120 |
+
"learning_rate": 4.301388235756442e-05,
|
| 7121 |
+
"loss": 2.1455,
|
| 7122 |
+
"step": 91100
|
| 7123 |
+
},
|
| 7124 |
+
{
|
| 7125 |
+
"epoch": 0.00613979760994424,
|
| 7126 |
+
"grad_norm": 1.6170133352279663,
|
| 7127 |
+
"learning_rate": 4.299852314097894e-05,
|
| 7128 |
+
"loss": 2.1379,
|
| 7129 |
+
"step": 91200
|
| 7130 |
+
},
|
| 7131 |
+
{
|
| 7132 |
+
"epoch": 0.0064188793194871595,
|
| 7133 |
+
"grad_norm": 1.5804738998413086,
|
| 7134 |
+
"learning_rate": 4.298314980776594e-05,
|
| 7135 |
+
"loss": 2.1415,
|
| 7136 |
+
"step": 91300
|
| 7137 |
+
},
|
| 7138 |
+
{
|
| 7139 |
+
"epoch": 0.006697961029030079,
|
| 7140 |
+
"grad_norm": 1.6626943349838257,
|
| 7141 |
+
"learning_rate": 4.2967762369983065e-05,
|
| 7142 |
+
"loss": 2.1513,
|
| 7143 |
+
"step": 91400
|
| 7144 |
+
},
|
| 7145 |
+
{
|
| 7146 |
+
"epoch": 0.006977042738573,
|
| 7147 |
+
"grad_norm": 1.5107390880584717,
|
| 7148 |
+
"learning_rate": 4.295236083969899e-05,
|
| 7149 |
+
"loss": 2.1477,
|
| 7150 |
+
"step": 91500
|
| 7151 |
+
},
|
| 7152 |
+
{
|
| 7153 |
+
"epoch": 0.0072561244481159195,
|
| 7154 |
+
"grad_norm": 1.6073352098464966,
|
| 7155 |
+
"learning_rate": 4.293694522899349e-05,
|
| 7156 |
+
"loss": 2.1468,
|
| 7157 |
+
"step": 91600
|
| 7158 |
+
},
|
| 7159 |
+
{
|
| 7160 |
+
"epoch": 0.007535206157658839,
|
| 7161 |
+
"grad_norm": 1.5642460584640503,
|
| 7162 |
+
"learning_rate": 4.292151554995734e-05,
|
| 7163 |
+
"loss": 2.1394,
|
| 7164 |
+
"step": 91700
|
| 7165 |
+
},
|
| 7166 |
+
{
|
| 7167 |
+
"epoch": 0.00781428786720176,
|
| 7168 |
+
"grad_norm": 1.4667211771011353,
|
| 7169 |
+
"learning_rate": 4.290607181469236e-05,
|
| 7170 |
+
"loss": 2.1446,
|
| 7171 |
+
"step": 91800
|
| 7172 |
+
},
|
| 7173 |
+
{
|
| 7174 |
+
"epoch": 0.00809336957674468,
|
| 7175 |
+
"grad_norm": 1.5254555940628052,
|
| 7176 |
+
"learning_rate": 4.2890614035311425e-05,
|
| 7177 |
+
"loss": 2.1516,
|
| 7178 |
+
"step": 91900
|
| 7179 |
+
},
|
| 7180 |
+
{
|
| 7181 |
+
"epoch": 0.008372451286287599,
|
| 7182 |
+
"grad_norm": 1.5663783550262451,
|
| 7183 |
+
"learning_rate": 4.2875142223938395e-05,
|
| 7184 |
+
"loss": 2.1407,
|
| 7185 |
+
"step": 92000
|
| 7186 |
+
},
|
| 7187 |
+
{
|
| 7188 |
+
"epoch": 0.008372451286287599,
|
| 7189 |
+
"eval_loss": 2.2744436264038086,
|
| 7190 |
+
"eval_runtime": 51.2413,
|
| 7191 |
+
"eval_samples_per_second": 198.941,
|
| 7192 |
+
"eval_steps_per_second": 1.561,
|
| 7193 |
+
"step": 92000
|
| 7194 |
+
},
|
| 7195 |
+
{
|
| 7196 |
+
"epoch": 0.008651532995830519,
|
| 7197 |
+
"grad_norm": 1.6381663084030151,
|
| 7198 |
+
"learning_rate": 4.285965639270814e-05,
|
| 7199 |
+
"loss": 2.1364,
|
| 7200 |
+
"step": 92100
|
| 7201 |
+
},
|
| 7202 |
+
{
|
| 7203 |
+
"epoch": 0.008930614705373438,
|
| 7204 |
+
"grad_norm": 1.5091722011566162,
|
| 7205 |
+
"learning_rate": 4.284415655376654e-05,
|
| 7206 |
+
"loss": 2.1387,
|
| 7207 |
+
"step": 92200
|
| 7208 |
+
},
|
| 7209 |
+
{
|
| 7210 |
+
"epoch": 0.00920969641491636,
|
| 7211 |
+
"grad_norm": 1.53826105594635,
|
| 7212 |
+
"learning_rate": 4.282864271927042e-05,
|
| 7213 |
+
"loss": 2.1422,
|
| 7214 |
+
"step": 92300
|
| 7215 |
+
},
|
| 7216 |
+
{
|
| 7217 |
+
"epoch": 0.00948877812445928,
|
| 7218 |
+
"grad_norm": 1.5313266515731812,
|
| 7219 |
+
"learning_rate": 4.281311490138765e-05,
|
| 7220 |
+
"loss": 2.1358,
|
| 7221 |
+
"step": 92400
|
| 7222 |
+
},
|
| 7223 |
+
{
|
| 7224 |
+
"epoch": 0.0097678598340022,
|
| 7225 |
+
"grad_norm": 1.5104668140411377,
|
| 7226 |
+
"learning_rate": 4.279757311229702e-05,
|
| 7227 |
+
"loss": 2.1324,
|
| 7228 |
+
"step": 92500
|
| 7229 |
+
},
|
| 7230 |
+
{
|
| 7231 |
+
"epoch": 0.010046941543545119,
|
| 7232 |
+
"grad_norm": 1.5808652639389038,
|
| 7233 |
+
"learning_rate": 4.278201736418828e-05,
|
| 7234 |
+
"loss": 2.1479,
|
| 7235 |
+
"step": 92600
|
| 7236 |
+
},
|
| 7237 |
+
{
|
| 7238 |
+
"epoch": 0.010326023253088039,
|
| 7239 |
+
"grad_norm": 1.5481529235839844,
|
| 7240 |
+
"learning_rate": 4.276644766926213e-05,
|
| 7241 |
+
"loss": 2.1359,
|
| 7242 |
+
"step": 92700
|
| 7243 |
+
},
|
| 7244 |
+
{
|
| 7245 |
+
"epoch": 0.01060510496263096,
|
| 7246 |
+
"grad_norm": 1.6417638063430786,
|
| 7247 |
+
"learning_rate": 4.275086403973023e-05,
|
| 7248 |
+
"loss": 2.1421,
|
| 7249 |
+
"step": 92800
|
| 7250 |
+
},
|
| 7251 |
+
{
|
| 7252 |
+
"epoch": 0.01088418667217388,
|
| 7253 |
+
"grad_norm": 1.6985423564910889,
|
| 7254 |
+
"learning_rate": 4.2735266487815156e-05,
|
| 7255 |
+
"loss": 2.1376,
|
| 7256 |
+
"step": 92900
|
| 7257 |
+
},
|
| 7258 |
+
{
|
| 7259 |
+
"epoch": 0.0111632683817168,
|
| 7260 |
+
"grad_norm": 1.7289358377456665,
|
| 7261 |
+
"learning_rate": 4.271965502575039e-05,
|
| 7262 |
+
"loss": 2.1446,
|
| 7263 |
+
"step": 93000
|
| 7264 |
+
},
|
| 7265 |
+
{
|
| 7266 |
+
"epoch": 0.0111632683817168,
|
| 7267 |
+
"eval_loss": 2.2753000259399414,
|
| 7268 |
+
"eval_runtime": 51.3478,
|
| 7269 |
+
"eval_samples_per_second": 198.529,
|
| 7270 |
+
"eval_steps_per_second": 1.558,
|
| 7271 |
+
"step": 93000
|
| 7272 |
+
},
|
| 7273 |
+
{
|
| 7274 |
+
"epoch": 0.011442350091259719,
|
| 7275 |
+
"grad_norm": 1.5933712720870972,
|
| 7276 |
+
"learning_rate": 4.2704029665780354e-05,
|
| 7277 |
+
"loss": 2.1341,
|
| 7278 |
+
"step": 93100
|
| 7279 |
+
},
|
| 7280 |
+
{
|
| 7281 |
+
"epoch": 0.011721431800802639,
|
| 7282 |
+
"grad_norm": 1.4210540056228638,
|
| 7283 |
+
"learning_rate": 4.2688390420160335e-05,
|
| 7284 |
+
"loss": 2.125,
|
| 7285 |
+
"step": 93200
|
| 7286 |
+
},
|
| 7287 |
+
{
|
| 7288 |
+
"epoch": 0.012000513510345558,
|
| 7289 |
+
"grad_norm": 1.7447474002838135,
|
| 7290 |
+
"learning_rate": 4.267273730115654e-05,
|
| 7291 |
+
"loss": 2.1259,
|
| 7292 |
+
"step": 93300
|
| 7293 |
+
},
|
| 7294 |
+
{
|
| 7295 |
+
"epoch": 0.01227959521988848,
|
| 7296 |
+
"grad_norm": 1.6314444541931152,
|
| 7297 |
+
"learning_rate": 4.265707032104603e-05,
|
| 7298 |
+
"loss": 2.1334,
|
| 7299 |
+
"step": 93400
|
| 7300 |
+
},
|
| 7301 |
+
{
|
| 7302 |
+
"epoch": 0.0125586769294314,
|
| 7303 |
+
"grad_norm": 1.6820199489593506,
|
| 7304 |
+
"learning_rate": 4.264138949211678e-05,
|
| 7305 |
+
"loss": 2.1263,
|
| 7306 |
+
"step": 93500
|
| 7307 |
+
},
|
| 7308 |
+
{
|
| 7309 |
+
"epoch": 0.012837758638974319,
|
| 7310 |
+
"grad_norm": 1.6244908571243286,
|
| 7311 |
+
"learning_rate": 4.2625694826667576e-05,
|
| 7312 |
+
"loss": 2.1313,
|
| 7313 |
+
"step": 93600
|
| 7314 |
+
},
|
| 7315 |
+
{
|
| 7316 |
+
"epoch": 0.013116840348517239,
|
| 7317 |
+
"grad_norm": 1.6561076641082764,
|
| 7318 |
+
"learning_rate": 4.260998633700809e-05,
|
| 7319 |
+
"loss": 2.1274,
|
| 7320 |
+
"step": 93700
|
| 7321 |
+
},
|
| 7322 |
+
{
|
| 7323 |
+
"epoch": 0.013395922058060158,
|
| 7324 |
+
"grad_norm": 1.5971171855926514,
|
| 7325 |
+
"learning_rate": 4.259426403545883e-05,
|
| 7326 |
+
"loss": 2.1304,
|
| 7327 |
+
"step": 93800
|
| 7328 |
+
},
|
| 7329 |
+
{
|
| 7330 |
+
"epoch": 0.013675003767603078,
|
| 7331 |
+
"grad_norm": 1.5423985719680786,
|
| 7332 |
+
"learning_rate": 4.257852793435113e-05,
|
| 7333 |
+
"loss": 2.1356,
|
| 7334 |
+
"step": 93900
|
| 7335 |
+
},
|
| 7336 |
+
{
|
| 7337 |
+
"epoch": 0.013954085477146,
|
| 7338 |
+
"grad_norm": 1.6720298528671265,
|
| 7339 |
+
"learning_rate": 4.256277804602715e-05,
|
| 7340 |
+
"loss": 2.1297,
|
| 7341 |
+
"step": 94000
|
| 7342 |
+
},
|
| 7343 |
+
{
|
| 7344 |
+
"epoch": 0.013954085477146,
|
| 7345 |
+
"eval_loss": 2.2692244052886963,
|
| 7346 |
+
"eval_runtime": 51.3308,
|
| 7347 |
+
"eval_samples_per_second": 198.594,
|
| 7348 |
+
"eval_steps_per_second": 1.559,
|
| 7349 |
+
"step": 94000
|
| 7350 |
+
},
|
| 7351 |
+
{
|
| 7352 |
+
"epoch": 0.01423316718668892,
|
| 7353 |
+
"grad_norm": 1.670507550239563,
|
| 7354 |
+
"learning_rate": 4.254701438283987e-05,
|
| 7355 |
+
"loss": 2.1326,
|
| 7356 |
+
"step": 94100
|
| 7357 |
+
},
|
| 7358 |
+
{
|
| 7359 |
+
"epoch": 0.014512248896231839,
|
| 7360 |
+
"grad_norm": 1.6407012939453125,
|
| 7361 |
+
"learning_rate": 4.253123695715307e-05,
|
| 7362 |
+
"loss": 2.1201,
|
| 7363 |
+
"step": 94200
|
| 7364 |
+
},
|
| 7365 |
+
{
|
| 7366 |
+
"epoch": 0.014791330605774759,
|
| 7367 |
+
"grad_norm": 1.6673415899276733,
|
| 7368 |
+
"learning_rate": 4.2515445781341306e-05,
|
| 7369 |
+
"loss": 2.1236,
|
| 7370 |
+
"step": 94300
|
| 7371 |
+
},
|
| 7372 |
+
{
|
| 7373 |
+
"epoch": 0.015070412315317678,
|
| 7374 |
+
"grad_norm": 1.6600122451782227,
|
| 7375 |
+
"learning_rate": 4.2499640867789955e-05,
|
| 7376 |
+
"loss": 2.1387,
|
| 7377 |
+
"step": 94400
|
| 7378 |
+
},
|
| 7379 |
+
{
|
| 7380 |
+
"epoch": 0.015349494024860598,
|
| 7381 |
+
"grad_norm": 1.669821858406067,
|
| 7382 |
+
"learning_rate": 4.248382222889515e-05,
|
| 7383 |
+
"loss": 2.1376,
|
| 7384 |
+
"step": 94500
|
| 7385 |
+
},
|
| 7386 |
+
{
|
| 7387 |
+
"epoch": 0.01562857573440352,
|
| 7388 |
+
"grad_norm": 1.6163934469223022,
|
| 7389 |
+
"learning_rate": 4.246798987706378e-05,
|
| 7390 |
+
"loss": 2.1219,
|
| 7391 |
+
"step": 94600
|
| 7392 |
+
},
|
| 7393 |
+
{
|
| 7394 |
+
"epoch": 0.015907657443946437,
|
| 7395 |
+
"grad_norm": 1.6449049711227417,
|
| 7396 |
+
"learning_rate": 4.24521438247135e-05,
|
| 7397 |
+
"loss": 2.1197,
|
| 7398 |
+
"step": 94700
|
| 7399 |
+
},
|
| 7400 |
+
{
|
| 7401 |
+
"epoch": 0.01618673915348936,
|
| 7402 |
+
"grad_norm": 1.6894525289535522,
|
| 7403 |
+
"learning_rate": 4.2436284084272706e-05,
|
| 7404 |
+
"loss": 2.1185,
|
| 7405 |
+
"step": 94800
|
| 7406 |
+
},
|
| 7407 |
+
{
|
| 7408 |
+
"epoch": 0.01646582086303228,
|
| 7409 |
+
"grad_norm": 1.6635791063308716,
|
| 7410 |
+
"learning_rate": 4.242041066818053e-05,
|
| 7411 |
+
"loss": 2.1202,
|
| 7412 |
+
"step": 94900
|
| 7413 |
+
},
|
| 7414 |
+
{
|
| 7415 |
+
"epoch": 0.016744902572575198,
|
| 7416 |
+
"grad_norm": 1.5680242776870728,
|
| 7417 |
+
"learning_rate": 4.240452358888685e-05,
|
| 7418 |
+
"loss": 2.1157,
|
| 7419 |
+
"step": 95000
|
| 7420 |
+
},
|
| 7421 |
+
{
|
| 7422 |
+
"epoch": 0.016744902572575198,
|
| 7423 |
+
"eval_loss": 2.2662878036499023,
|
| 7424 |
+
"eval_runtime": 51.3865,
|
| 7425 |
+
"eval_samples_per_second": 198.379,
|
| 7426 |
+
"eval_steps_per_second": 1.557,
|
| 7427 |
+
"step": 95000
|
| 7428 |
+
},
|
| 7429 |
+
{
|
| 7430 |
+
"epoch": 0.01702398428211812,
|
| 7431 |
+
"grad_norm": 1.6089997291564941,
|
| 7432 |
+
"learning_rate": 4.2388622858852224e-05,
|
| 7433 |
+
"loss": 2.1249,
|
| 7434 |
+
"step": 95100
|
| 7435 |
+
},
|
| 7436 |
+
{
|
| 7437 |
+
"epoch": 0.017303065991661037,
|
| 7438 |
+
"grad_norm": 1.5722119808197021,
|
| 7439 |
+
"learning_rate": 4.237270849054794e-05,
|
| 7440 |
+
"loss": 2.1169,
|
| 7441 |
+
"step": 95200
|
| 7442 |
+
},
|
| 7443 |
+
{
|
| 7444 |
+
"epoch": 0.01758214770120396,
|
| 7445 |
+
"grad_norm": 1.6253992319107056,
|
| 7446 |
+
"learning_rate": 4.2356780496455984e-05,
|
| 7447 |
+
"loss": 2.121,
|
| 7448 |
+
"step": 95300
|
| 7449 |
+
},
|
| 7450 |
+
{
|
| 7451 |
+
"epoch": 0.017861229410746877,
|
| 7452 |
+
"grad_norm": 1.7267462015151978,
|
| 7453 |
+
"learning_rate": 4.2340838889069014e-05,
|
| 7454 |
+
"loss": 2.1174,
|
| 7455 |
+
"step": 95400
|
| 7456 |
+
},
|
| 7457 |
+
{
|
| 7458 |
+
"epoch": 0.018140311120289798,
|
| 7459 |
+
"grad_norm": 1.639393925666809,
|
| 7460 |
+
"learning_rate": 4.232488368089038e-05,
|
| 7461 |
+
"loss": 2.1191,
|
| 7462 |
+
"step": 95500
|
| 7463 |
+
},
|
| 7464 |
+
{
|
| 7465 |
+
"epoch": 0.01841939282983272,
|
| 7466 |
+
"grad_norm": 1.570025086402893,
|
| 7467 |
+
"learning_rate": 4.2308914884434096e-05,
|
| 7468 |
+
"loss": 2.1208,
|
| 7469 |
+
"step": 95600
|
| 7470 |
+
},
|
| 7471 |
+
{
|
| 7472 |
+
"epoch": 0.018698474539375638,
|
| 7473 |
+
"grad_norm": 1.6924635171890259,
|
| 7474 |
+
"learning_rate": 4.2292932512224835e-05,
|
| 7475 |
+
"loss": 2.1218,
|
| 7476 |
+
"step": 95700
|
| 7477 |
+
},
|
| 7478 |
+
{
|
| 7479 |
+
"epoch": 0.01897755624891856,
|
| 7480 |
+
"grad_norm": 1.5708248615264893,
|
| 7481 |
+
"learning_rate": 4.22769365767979e-05,
|
| 7482 |
+
"loss": 2.1191,
|
| 7483 |
+
"step": 95800
|
| 7484 |
+
},
|
| 7485 |
+
{
|
| 7486 |
+
"epoch": 0.019256637958461477,
|
| 7487 |
+
"grad_norm": 1.6411412954330444,
|
| 7488 |
+
"learning_rate": 4.226092709069926e-05,
|
| 7489 |
+
"loss": 2.1017,
|
| 7490 |
+
"step": 95900
|
| 7491 |
+
},
|
| 7492 |
+
{
|
| 7493 |
+
"epoch": 0.0195357196680044,
|
| 7494 |
+
"grad_norm": 1.6997929811477661,
|
| 7495 |
+
"learning_rate": 4.224490406648548e-05,
|
| 7496 |
+
"loss": 2.1125,
|
| 7497 |
+
"step": 96000
|
| 7498 |
+
},
|
| 7499 |
+
{
|
| 7500 |
+
"epoch": 0.0195357196680044,
|
| 7501 |
+
"eval_loss": 2.2651915550231934,
|
| 7502 |
+
"eval_runtime": 51.6612,
|
| 7503 |
+
"eval_samples_per_second": 197.324,
|
| 7504 |
+
"eval_steps_per_second": 1.549,
|
| 7505 |
+
"step": 96000
|
| 7506 |
+
},
|
| 7507 |
+
{
|
| 7508 |
+
"epoch": 0.01981480137754732,
|
| 7509 |
+
"grad_norm": 1.678976058959961,
|
| 7510 |
+
"learning_rate": 4.222886751672379e-05,
|
| 7511 |
+
"loss": 2.1085,
|
| 7512 |
+
"step": 96100
|
| 7513 |
+
},
|
| 7514 |
+
{
|
| 7515 |
+
"epoch": 0.020093883087090238,
|
| 7516 |
+
"grad_norm": 1.6502765417099,
|
| 7517 |
+
"learning_rate": 4.221281745399197e-05,
|
| 7518 |
+
"loss": 2.104,
|
| 7519 |
+
"step": 96200
|
| 7520 |
+
},
|
| 7521 |
+
{
|
| 7522 |
+
"epoch": 0.02037296479663316,
|
| 7523 |
+
"grad_norm": 1.7067300081253052,
|
| 7524 |
+
"learning_rate": 4.219675389087845e-05,
|
| 7525 |
+
"loss": 2.0918,
|
| 7526 |
+
"step": 96300
|
| 7527 |
+
},
|
| 7528 |
+
{
|
| 7529 |
+
"epoch": 0.020652046506176077,
|
| 7530 |
+
"grad_norm": 1.5428402423858643,
|
| 7531 |
+
"learning_rate": 4.218067683998221e-05,
|
| 7532 |
+
"loss": 2.0912,
|
| 7533 |
+
"step": 96400
|
| 7534 |
+
},
|
| 7535 |
+
{
|
| 7536 |
+
"epoch": 0.020931128215719,
|
| 7537 |
+
"grad_norm": 1.5912175178527832,
|
| 7538 |
+
"learning_rate": 4.2164586313912844e-05,
|
| 7539 |
+
"loss": 2.1018,
|
| 7540 |
+
"step": 96500
|
| 7541 |
+
},
|
| 7542 |
+
{
|
| 7543 |
+
"epoch": 0.02121020992526192,
|
| 7544 |
+
"grad_norm": 1.6541246175765991,
|
| 7545 |
+
"learning_rate": 4.214848232529048e-05,
|
| 7546 |
+
"loss": 2.1059,
|
| 7547 |
+
"step": 96600
|
| 7548 |
+
},
|
| 7549 |
+
{
|
| 7550 |
+
"epoch": 0.021489291634804838,
|
| 7551 |
+
"grad_norm": 1.6385318040847778,
|
| 7552 |
+
"learning_rate": 4.2132364886745834e-05,
|
| 7553 |
+
"loss": 2.0992,
|
| 7554 |
+
"step": 96700
|
| 7555 |
+
},
|
| 7556 |
+
{
|
| 7557 |
+
"epoch": 0.02176837334434776,
|
| 7558 |
+
"grad_norm": 1.6574422121047974,
|
| 7559 |
+
"learning_rate": 4.2116234010920153e-05,
|
| 7560 |
+
"loss": 2.1014,
|
| 7561 |
+
"step": 96800
|
| 7562 |
+
},
|
| 7563 |
+
{
|
| 7564 |
+
"epoch": 0.022047455053890677,
|
| 7565 |
+
"grad_norm": 1.798996090888977,
|
| 7566 |
+
"learning_rate": 4.210008971046522e-05,
|
| 7567 |
+
"loss": 2.0837,
|
| 7568 |
+
"step": 96900
|
| 7569 |
+
},
|
| 7570 |
+
{
|
| 7571 |
+
"epoch": 0.0223265367634336,
|
| 7572 |
+
"grad_norm": 1.7654832601547241,
|
| 7573 |
+
"learning_rate": 4.208393199804337e-05,
|
| 7574 |
+
"loss": 2.0967,
|
| 7575 |
+
"step": 97000
|
| 7576 |
+
},
|
| 7577 |
+
{
|
| 7578 |
+
"epoch": 0.0223265367634336,
|
| 7579 |
+
"eval_loss": 2.2732906341552734,
|
| 7580 |
+
"eval_runtime": 51.5925,
|
| 7581 |
+
"eval_samples_per_second": 197.587,
|
| 7582 |
+
"eval_steps_per_second": 1.551,
|
| 7583 |
+
"step": 97000
|
| 7584 |
+
},
|
| 7585 |
+
{
|
| 7586 |
+
"epoch": 0.022605618472976517,
|
| 7587 |
+
"grad_norm": 1.659149169921875,
|
| 7588 |
+
"learning_rate": 4.206776088632744e-05,
|
| 7589 |
+
"loss": 2.0867,
|
| 7590 |
+
"step": 97100
|
| 7591 |
+
},
|
| 7592 |
+
{
|
| 7593 |
+
"epoch": 0.022884700182519438,
|
| 7594 |
+
"grad_norm": 1.6352757215499878,
|
| 7595 |
+
"learning_rate": 4.205157638800077e-05,
|
| 7596 |
+
"loss": 2.0883,
|
| 7597 |
+
"step": 97200
|
| 7598 |
+
},
|
| 7599 |
+
{
|
| 7600 |
+
"epoch": 0.02316378189206236,
|
| 7601 |
+
"grad_norm": 1.6729434728622437,
|
| 7602 |
+
"learning_rate": 4.203537851575722e-05,
|
| 7603 |
+
"loss": 2.0826,
|
| 7604 |
+
"step": 97300
|
| 7605 |
+
},
|
| 7606 |
+
{
|
| 7607 |
+
"epoch": 0.023442863601605277,
|
| 7608 |
+
"grad_norm": 1.6211354732513428,
|
| 7609 |
+
"learning_rate": 4.201916728230112e-05,
|
| 7610 |
+
"loss": 2.0862,
|
| 7611 |
+
"step": 97400
|
| 7612 |
+
},
|
| 7613 |
+
{
|
| 7614 |
+
"epoch": 0.0237219453111482,
|
| 7615 |
+
"grad_norm": 1.6426453590393066,
|
| 7616 |
+
"learning_rate": 4.20029427003473e-05,
|
| 7617 |
+
"loss": 2.1016,
|
| 7618 |
+
"step": 97500
|
| 7619 |
+
},
|
| 7620 |
+
{
|
| 7621 |
+
"epoch": 0.024001027020691117,
|
| 7622 |
+
"grad_norm": 1.7282731533050537,
|
| 7623 |
+
"learning_rate": 4.198670478262103e-05,
|
| 7624 |
+
"loss": 2.0897,
|
| 7625 |
+
"step": 97600
|
| 7626 |
+
},
|
| 7627 |
+
{
|
| 7628 |
+
"epoch": 0.024280108730234038,
|
| 7629 |
+
"grad_norm": 1.6720244884490967,
|
| 7630 |
+
"learning_rate": 4.1970453541858075e-05,
|
| 7631 |
+
"loss": 2.0887,
|
| 7632 |
+
"step": 97700
|
| 7633 |
+
},
|
| 7634 |
+
{
|
| 7635 |
+
"epoch": 0.02455919043977696,
|
| 7636 |
+
"grad_norm": 1.6182509660720825,
|
| 7637 |
+
"learning_rate": 4.195418899080462e-05,
|
| 7638 |
+
"loss": 2.0941,
|
| 7639 |
+
"step": 97800
|
| 7640 |
+
},
|
| 7641 |
+
{
|
| 7642 |
+
"epoch": 0.024838272149319877,
|
| 7643 |
+
"grad_norm": 1.5953019857406616,
|
| 7644 |
+
"learning_rate": 4.19379111422173e-05,
|
| 7645 |
+
"loss": 2.0799,
|
| 7646 |
+
"step": 97900
|
| 7647 |
+
},
|
| 7648 |
+
{
|
| 7649 |
+
"epoch": 0.0251173538588628,
|
| 7650 |
+
"grad_norm": 1.6194970607757568,
|
| 7651 |
+
"learning_rate": 4.1921620008863193e-05,
|
| 7652 |
+
"loss": 2.0915,
|
| 7653 |
+
"step": 98000
|
| 7654 |
+
},
|
| 7655 |
+
{
|
| 7656 |
+
"epoch": 0.0251173538588628,
|
| 7657 |
+
"eval_loss": 2.269259452819824,
|
| 7658 |
+
"eval_runtime": 51.5672,
|
| 7659 |
+
"eval_samples_per_second": 197.684,
|
| 7660 |
+
"eval_steps_per_second": 1.551,
|
| 7661 |
+
"step": 98000
|
| 7662 |
+
},
|
| 7663 |
+
{
|
| 7664 |
+
"epoch": 0.025396435568405717,
|
| 7665 |
+
"grad_norm": 1.6391347646713257,
|
| 7666 |
+
"learning_rate": 4.1905315603519765e-05,
|
| 7667 |
+
"loss": 2.0761,
|
| 7668 |
+
"step": 98100
|
| 7669 |
+
},
|
| 7670 |
+
{
|
| 7671 |
+
"epoch": 0.025675517277948638,
|
| 7672 |
+
"grad_norm": 1.6369253396987915,
|
| 7673 |
+
"learning_rate": 4.1888997938974935e-05,
|
| 7674 |
+
"loss": 2.0862,
|
| 7675 |
+
"step": 98200
|
| 7676 |
+
},
|
| 7677 |
+
{
|
| 7678 |
+
"epoch": 0.025954598987491556,
|
| 7679 |
+
"grad_norm": 1.628437876701355,
|
| 7680 |
+
"learning_rate": 4.187266702802698e-05,
|
| 7681 |
+
"loss": 2.0835,
|
| 7682 |
+
"step": 98300
|
| 7683 |
+
},
|
| 7684 |
+
{
|
| 7685 |
+
"epoch": 0.026233680697034478,
|
| 7686 |
+
"grad_norm": 1.643228530883789,
|
| 7687 |
+
"learning_rate": 4.1856322883484584e-05,
|
| 7688 |
+
"loss": 2.078,
|
| 7689 |
+
"step": 98400
|
| 7690 |
+
},
|
| 7691 |
+
{
|
| 7692 |
+
"epoch": 0.0265127624065774,
|
| 7693 |
+
"grad_norm": 1.6307592391967773,
|
| 7694 |
+
"learning_rate": 4.183996551816681e-05,
|
| 7695 |
+
"loss": 2.0703,
|
| 7696 |
+
"step": 98500
|
| 7697 |
+
},
|
| 7698 |
+
{
|
| 7699 |
+
"epoch": 0.026791844116120317,
|
| 7700 |
+
"grad_norm": 1.6314188241958618,
|
| 7701 |
+
"learning_rate": 4.18235949449031e-05,
|
| 7702 |
+
"loss": 2.0926,
|
| 7703 |
+
"step": 98600
|
| 7704 |
+
},
|
| 7705 |
+
{
|
| 7706 |
+
"epoch": 0.02707092582566324,
|
| 7707 |
+
"grad_norm": 1.5619598627090454,
|
| 7708 |
+
"learning_rate": 4.180721117653323e-05,
|
| 7709 |
+
"loss": 2.0827,
|
| 7710 |
+
"step": 98700
|
| 7711 |
+
},
|
| 7712 |
+
{
|
| 7713 |
+
"epoch": 0.027350007535206156,
|
| 7714 |
+
"grad_norm": 1.5566825866699219,
|
| 7715 |
+
"learning_rate": 4.179081422590736e-05,
|
| 7716 |
+
"loss": 2.0754,
|
| 7717 |
+
"step": 98800
|
| 7718 |
+
},
|
| 7719 |
+
{
|
| 7720 |
+
"epoch": 0.027629089244749078,
|
| 7721 |
+
"grad_norm": 1.621233344078064,
|
| 7722 |
+
"learning_rate": 4.177440410588596e-05,
|
| 7723 |
+
"loss": 2.0702,
|
| 7724 |
+
"step": 98900
|
| 7725 |
+
},
|
| 7726 |
+
{
|
| 7727 |
+
"epoch": 0.027908170954292,
|
| 7728 |
+
"grad_norm": 1.6783591508865356,
|
| 7729 |
+
"learning_rate": 4.1757980829339826e-05,
|
| 7730 |
+
"loss": 2.0691,
|
| 7731 |
+
"step": 99000
|
| 7732 |
+
},
|
| 7733 |
+
{
|
| 7734 |
+
"epoch": 0.027908170954292,
|
| 7735 |
+
"eval_loss": 2.2590763568878174,
|
| 7736 |
+
"eval_runtime": 51.6452,
|
| 7737 |
+
"eval_samples_per_second": 197.385,
|
| 7738 |
+
"eval_steps_per_second": 1.549,
|
| 7739 |
+
"step": 99000
|
| 7740 |
+
},
|
| 7741 |
+
{
|
| 7742 |
+
"epoch": 0.028187252663834917,
|
| 7743 |
+
"grad_norm": 1.6113650798797607,
|
| 7744 |
+
"learning_rate": 4.1741544409150104e-05,
|
| 7745 |
+
"loss": 2.0677,
|
| 7746 |
+
"step": 99100
|
| 7747 |
+
},
|
| 7748 |
+
{
|
| 7749 |
+
"epoch": 0.02846633437337784,
|
| 7750 |
+
"grad_norm": 1.6120631694793701,
|
| 7751 |
+
"learning_rate": 4.172509485820823e-05,
|
| 7752 |
+
"loss": 2.0784,
|
| 7753 |
+
"step": 99200
|
| 7754 |
+
},
|
| 7755 |
+
{
|
| 7756 |
+
"epoch": 0.028745416082920756,
|
| 7757 |
+
"grad_norm": 1.603555679321289,
|
| 7758 |
+
"learning_rate": 4.170863218941593e-05,
|
| 7759 |
+
"loss": 2.0685,
|
| 7760 |
+
"step": 99300
|
| 7761 |
+
},
|
| 7762 |
+
{
|
| 7763 |
+
"epoch": 0.029024497792463678,
|
| 7764 |
+
"grad_norm": 1.5876059532165527,
|
| 7765 |
+
"learning_rate": 4.1692156415685234e-05,
|
| 7766 |
+
"loss": 2.0694,
|
| 7767 |
+
"step": 99400
|
| 7768 |
+
},
|
| 7769 |
+
{
|
| 7770 |
+
"epoch": 0.0293035795020066,
|
| 7771 |
+
"grad_norm": 1.5650913715362549,
|
| 7772 |
+
"learning_rate": 4.167566754993844e-05,
|
| 7773 |
+
"loss": 2.0714,
|
| 7774 |
+
"step": 99500
|
| 7775 |
+
},
|
| 7776 |
+
{
|
| 7777 |
+
"epoch": 0.029582661211549517,
|
| 7778 |
+
"grad_norm": 1.6612671613693237,
|
| 7779 |
+
"learning_rate": 4.1659165605108134e-05,
|
| 7780 |
+
"loss": 2.0754,
|
| 7781 |
+
"step": 99600
|
| 7782 |
+
},
|
| 7783 |
+
{
|
| 7784 |
+
"epoch": 0.02986174292109244,
|
| 7785 |
+
"grad_norm": 1.6820577383041382,
|
| 7786 |
+
"learning_rate": 4.1642650594137116e-05,
|
| 7787 |
+
"loss": 2.0686,
|
| 7788 |
+
"step": 99700
|
| 7789 |
+
},
|
| 7790 |
+
{
|
| 7791 |
+
"epoch": 0.030140824630635357,
|
| 7792 |
+
"grad_norm": 1.6811972856521606,
|
| 7793 |
+
"learning_rate": 4.162612252997849e-05,
|
| 7794 |
+
"loss": 2.0719,
|
| 7795 |
+
"step": 99800
|
| 7796 |
+
},
|
| 7797 |
+
{
|
| 7798 |
+
"epoch": 0.030419906340178278,
|
| 7799 |
+
"grad_norm": 1.6226907968521118,
|
| 7800 |
+
"learning_rate": 4.160958142559556e-05,
|
| 7801 |
+
"loss": 2.0654,
|
| 7802 |
+
"step": 99900
|
| 7803 |
+
},
|
| 7804 |
+
{
|
| 7805 |
+
"epoch": 0.030698988049721196,
|
| 7806 |
+
"grad_norm": 1.5367672443389893,
|
| 7807 |
+
"learning_rate": 4.159302729396186e-05,
|
| 7808 |
+
"loss": 2.077,
|
| 7809 |
+
"step": 100000
|
| 7810 |
+
},
|
| 7811 |
+
{
|
| 7812 |
+
"epoch": 0.030698988049721196,
|
| 7813 |
+
"eval_loss": 2.2724859714508057,
|
| 7814 |
+
"eval_runtime": 51.6394,
|
| 7815 |
+
"eval_samples_per_second": 197.407,
|
| 7816 |
+
"eval_steps_per_second": 1.549,
|
| 7817 |
+
"step": 100000
|
| 7818 |
+
},
|
| 7819 |
+
{
|
| 7820 |
+
"epoch": 0.030978069759264117,
|
| 7821 |
+
"grad_norm": 1.6159389019012451,
|
| 7822 |
+
"learning_rate": 4.157646014806117e-05,
|
| 7823 |
+
"loss": 2.0694,
|
| 7824 |
+
"step": 100100
|
| 7825 |
+
},
|
| 7826 |
+
{
|
| 7827 |
+
"epoch": 0.03125715146880704,
|
| 7828 |
+
"grad_norm": 1.6105570793151855,
|
| 7829 |
+
"learning_rate": 4.155988000088745e-05,
|
| 7830 |
+
"loss": 2.0527,
|
| 7831 |
+
"step": 100200
|
| 7832 |
+
},
|
| 7833 |
+
{
|
| 7834 |
+
"epoch": 0.03153623317834996,
|
| 7835 |
+
"grad_norm": 1.6409541368484497,
|
| 7836 |
+
"learning_rate": 4.1543286865444856e-05,
|
| 7837 |
+
"loss": 2.0671,
|
| 7838 |
+
"step": 100300
|
| 7839 |
+
},
|
| 7840 |
+
{
|
| 7841 |
+
"epoch": 0.031815314887892875,
|
| 7842 |
+
"grad_norm": 1.6089441776275635,
|
| 7843 |
+
"learning_rate": 4.152668075474775e-05,
|
| 7844 |
+
"loss": 2.0659,
|
| 7845 |
+
"step": 100400
|
| 7846 |
+
},
|
| 7847 |
+
{
|
| 7848 |
+
"epoch": 0.0320943965974358,
|
| 7849 |
+
"grad_norm": 1.6299337148666382,
|
| 7850 |
+
"learning_rate": 4.151006168182065e-05,
|
| 7851 |
+
"loss": 2.0634,
|
| 7852 |
+
"step": 100500
|
| 7853 |
+
},
|
| 7854 |
+
{
|
| 7855 |
+
"epoch": 0.03237347830697872,
|
| 7856 |
+
"grad_norm": 1.5447492599487305,
|
| 7857 |
+
"learning_rate": 4.1493429659698266e-05,
|
| 7858 |
+
"loss": 2.0709,
|
| 7859 |
+
"step": 100600
|
| 7860 |
+
},
|
| 7861 |
+
{
|
| 7862 |
+
"epoch": 0.032652560016521635,
|
| 7863 |
+
"grad_norm": 1.6599881649017334,
|
| 7864 |
+
"learning_rate": 4.147678470142544e-05,
|
| 7865 |
+
"loss": 2.0663,
|
| 7866 |
+
"step": 100700
|
| 7867 |
+
},
|
| 7868 |
+
{
|
| 7869 |
+
"epoch": 0.03293164172606456,
|
| 7870 |
+
"grad_norm": 1.6231052875518799,
|
| 7871 |
+
"learning_rate": 4.146012682005717e-05,
|
| 7872 |
+
"loss": 2.0691,
|
| 7873 |
+
"step": 100800
|
| 7874 |
+
},
|
| 7875 |
+
{
|
| 7876 |
+
"epoch": 0.03321072343560748,
|
| 7877 |
+
"grad_norm": 1.67854642868042,
|
| 7878 |
+
"learning_rate": 4.144345602865859e-05,
|
| 7879 |
+
"loss": 2.0569,
|
| 7880 |
+
"step": 100900
|
| 7881 |
+
},
|
| 7882 |
+
{
|
| 7883 |
+
"epoch": 0.033489805145150396,
|
| 7884 |
+
"grad_norm": 1.655781626701355,
|
| 7885 |
+
"learning_rate": 4.1426772340304964e-05,
|
| 7886 |
+
"loss": 2.0546,
|
| 7887 |
+
"step": 101000
|
| 7888 |
+
},
|
| 7889 |
+
{
|
| 7890 |
+
"epoch": 0.033489805145150396,
|
| 7891 |
+
"eval_loss": 2.275222063064575,
|
| 7892 |
+
"eval_runtime": 51.7111,
|
| 7893 |
+
"eval_samples_per_second": 197.134,
|
| 7894 |
+
"eval_steps_per_second": 1.547,
|
| 7895 |
+
"step": 101000
|
| 7896 |
+
},
|
| 7897 |
+
{
|
| 7898 |
+
"epoch": 0.033768886854693314,
|
| 7899 |
+
"grad_norm": 1.6492592096328735,
|
| 7900 |
+
"learning_rate": 4.141007576808166e-05,
|
| 7901 |
+
"loss": 2.0605,
|
| 7902 |
+
"step": 101100
|
| 7903 |
+
},
|
| 7904 |
+
{
|
| 7905 |
+
"epoch": 0.03404796856423624,
|
| 7906 |
+
"grad_norm": 1.5914729833602905,
|
| 7907 |
+
"learning_rate": 4.139336632508415e-05,
|
| 7908 |
+
"loss": 2.0417,
|
| 7909 |
+
"step": 101200
|
| 7910 |
+
},
|
| 7911 |
+
{
|
| 7912 |
+
"epoch": 0.03432705027377916,
|
| 7913 |
+
"grad_norm": 1.6000899076461792,
|
| 7914 |
+
"learning_rate": 4.1376644024418035e-05,
|
| 7915 |
+
"loss": 2.0621,
|
| 7916 |
+
"step": 101300
|
| 7917 |
+
},
|
| 7918 |
+
{
|
| 7919 |
+
"epoch": 0.034606131983322075,
|
| 7920 |
+
"grad_norm": 1.5423117876052856,
|
| 7921 |
+
"learning_rate": 4.135990887919894e-05,
|
| 7922 |
+
"loss": 2.0709,
|
| 7923 |
+
"step": 101400
|
| 7924 |
+
},
|
| 7925 |
+
{
|
| 7926 |
+
"epoch": 0.034885213692865,
|
| 7927 |
+
"grad_norm": 1.5954508781433105,
|
| 7928 |
+
"learning_rate": 4.134316090255263e-05,
|
| 7929 |
+
"loss": 2.0585,
|
| 7930 |
+
"step": 101500
|
| 7931 |
+
},
|
| 7932 |
+
{
|
| 7933 |
+
"epoch": 0.03516429540240792,
|
| 7934 |
+
"grad_norm": 1.603651523590088,
|
| 7935 |
+
"learning_rate": 4.1326400107614877e-05,
|
| 7936 |
+
"loss": 2.0688,
|
| 7937 |
+
"step": 101600
|
| 7938 |
+
},
|
| 7939 |
+
{
|
| 7940 |
+
"epoch": 0.035443377111950836,
|
| 7941 |
+
"grad_norm": 1.6470582485198975,
|
| 7942 |
+
"learning_rate": 4.130962650753154e-05,
|
| 7943 |
+
"loss": 2.0548,
|
| 7944 |
+
"step": 101700
|
| 7945 |
+
},
|
| 7946 |
+
{
|
| 7947 |
+
"epoch": 0.035722458821493754,
|
| 7948 |
+
"grad_norm": 1.6013911962509155,
|
| 7949 |
+
"learning_rate": 4.129284011545852e-05,
|
| 7950 |
+
"loss": 2.0502,
|
| 7951 |
+
"step": 101800
|
| 7952 |
+
},
|
| 7953 |
+
{
|
| 7954 |
+
"epoch": 0.03600154053103668,
|
| 7955 |
+
"grad_norm": 1.6595914363861084,
|
| 7956 |
+
"learning_rate": 4.127604094456174e-05,
|
| 7957 |
+
"loss": 2.0653,
|
| 7958 |
+
"step": 101900
|
| 7959 |
+
},
|
| 7960 |
+
{
|
| 7961 |
+
"epoch": 0.036280622240579596,
|
| 7962 |
+
"grad_norm": 1.6441375017166138,
|
| 7963 |
+
"learning_rate": 4.125922900801715e-05,
|
| 7964 |
+
"loss": 2.0181,
|
| 7965 |
+
"step": 102000
|
| 7966 |
+
},
|
| 7967 |
+
{
|
| 7968 |
+
"epoch": 0.036280622240579596,
|
| 7969 |
+
"eval_loss": 2.27272891998291,
|
| 7970 |
+
"eval_runtime": 51.6797,
|
| 7971 |
+
"eval_samples_per_second": 197.253,
|
| 7972 |
+
"eval_steps_per_second": 1.548,
|
| 7973 |
+
"step": 102000
|
| 7974 |
+
},
|
| 7975 |
+
{
|
| 7976 |
+
"epoch": 0.036559703950122514,
|
| 7977 |
+
"grad_norm": 1.5254052877426147,
|
| 7978 |
+
"learning_rate": 4.124240431901071e-05,
|
| 7979 |
+
"loss": 2.0585,
|
| 7980 |
+
"step": 102100
|
| 7981 |
+
},
|
| 7982 |
+
{
|
| 7983 |
+
"epoch": 0.03683878565966544,
|
| 7984 |
+
"grad_norm": 1.5843355655670166,
|
| 7985 |
+
"learning_rate": 4.1225566890738384e-05,
|
| 7986 |
+
"loss": 2.059,
|
| 7987 |
+
"step": 102200
|
| 7988 |
+
},
|
| 7989 |
+
{
|
| 7990 |
+
"epoch": 0.03711786736920836,
|
| 7991 |
+
"grad_norm": 1.64484441280365,
|
| 7992 |
+
"learning_rate": 4.120871673640613e-05,
|
| 7993 |
+
"loss": 2.0522,
|
| 7994 |
+
"step": 102300
|
| 7995 |
+
},
|
| 7996 |
+
{
|
| 7997 |
+
"epoch": 0.037396949078751275,
|
| 7998 |
+
"grad_norm": 1.7020013332366943,
|
| 7999 |
+
"learning_rate": 4.119185386922988e-05,
|
| 8000 |
+
"loss": 2.0519,
|
| 8001 |
+
"step": 102400
|
| 8002 |
+
},
|
| 8003 |
+
{
|
| 8004 |
+
"epoch": 0.0376760307882942,
|
| 8005 |
+
"grad_norm": 1.6111907958984375,
|
| 8006 |
+
"learning_rate": 4.117497830243555e-05,
|
| 8007 |
+
"loss": 2.0563,
|
| 8008 |
+
"step": 102500
|
| 8009 |
+
},
|
| 8010 |
+
{
|
| 8011 |
+
"epoch": 0.03795511249783712,
|
| 8012 |
+
"grad_norm": 1.5560317039489746,
|
| 8013 |
+
"learning_rate": 4.1158090049259005e-05,
|
| 8014 |
+
"loss": 2.0476,
|
| 8015 |
+
"step": 102600
|
| 8016 |
+
},
|
| 8017 |
+
{
|
| 8018 |
+
"epoch": 0.038234194207380036,
|
| 8019 |
+
"grad_norm": 1.6740190982818604,
|
| 8020 |
+
"learning_rate": 4.114118912294607e-05,
|
| 8021 |
+
"loss": 2.0654,
|
| 8022 |
+
"step": 102700
|
| 8023 |
+
},
|
| 8024 |
+
{
|
| 8025 |
+
"epoch": 0.038513275916922954,
|
| 8026 |
+
"grad_norm": 1.5790201425552368,
|
| 8027 |
+
"learning_rate": 4.1124275536752494e-05,
|
| 8028 |
+
"loss": 2.0557,
|
| 8029 |
+
"step": 102800
|
| 8030 |
+
},
|
| 8031 |
+
{
|
| 8032 |
+
"epoch": 0.03879235762646588,
|
| 8033 |
+
"grad_norm": 1.6047033071517944,
|
| 8034 |
+
"learning_rate": 4.110734930394397e-05,
|
| 8035 |
+
"loss": 2.0472,
|
| 8036 |
+
"step": 102900
|
| 8037 |
+
},
|
| 8038 |
+
{
|
| 8039 |
+
"epoch": 0.0390714393360088,
|
| 8040 |
+
"grad_norm": 1.5579067468643188,
|
| 8041 |
+
"learning_rate": 4.1090410437796104e-05,
|
| 8042 |
+
"loss": 2.054,
|
| 8043 |
+
"step": 103000
|
| 8044 |
+
},
|
| 8045 |
+
{
|
| 8046 |
+
"epoch": 0.0390714393360088,
|
| 8047 |
+
"eval_loss": 2.261636257171631,
|
| 8048 |
+
"eval_runtime": 51.6907,
|
| 8049 |
+
"eval_samples_per_second": 197.212,
|
| 8050 |
+
"eval_steps_per_second": 1.548,
|
| 8051 |
+
"step": 103000
|
| 8052 |
+
},
|
| 8053 |
+
{
|
| 8054 |
+
"epoch": 0.039350521045551715,
|
| 8055 |
+
"grad_norm": 1.5974478721618652,
|
| 8056 |
+
"learning_rate": 4.107345895159441e-05,
|
| 8057 |
+
"loss": 2.0528,
|
| 8058 |
+
"step": 103100
|
| 8059 |
+
},
|
| 8060 |
+
{
|
| 8061 |
+
"epoch": 0.03962960275509464,
|
| 8062 |
+
"grad_norm": 1.6119569540023804,
|
| 8063 |
+
"learning_rate": 4.105649485863431e-05,
|
| 8064 |
+
"loss": 2.0571,
|
| 8065 |
+
"step": 103200
|
| 8066 |
+
},
|
| 8067 |
+
{
|
| 8068 |
+
"epoch": 0.03990868446463756,
|
| 8069 |
+
"grad_norm": 1.606919527053833,
|
| 8070 |
+
"learning_rate": 4.1039518172221105e-05,
|
| 8071 |
+
"loss": 2.0516,
|
| 8072 |
+
"step": 103300
|
| 8073 |
+
},
|
| 8074 |
+
{
|
| 8075 |
+
"epoch": 0.040187766174180475,
|
| 8076 |
+
"grad_norm": 1.700379490852356,
|
| 8077 |
+
"learning_rate": 4.1022528905669954e-05,
|
| 8078 |
+
"loss": 2.0405,
|
| 8079 |
+
"step": 103400
|
| 8080 |
+
},
|
| 8081 |
+
{
|
| 8082 |
+
"epoch": 0.04046684788372339,
|
| 8083 |
+
"grad_norm": 1.5789330005645752,
|
| 8084 |
+
"learning_rate": 4.100552707230593e-05,
|
| 8085 |
+
"loss": 2.0551,
|
| 8086 |
+
"step": 103500
|
| 8087 |
+
},
|
| 8088 |
+
{
|
| 8089 |
+
"epoch": 0.04074592959326632,
|
| 8090 |
+
"grad_norm": 1.5743343830108643,
|
| 8091 |
+
"learning_rate": 4.098851268546392e-05,
|
| 8092 |
+
"loss": 2.0558,
|
| 8093 |
+
"step": 103600
|
| 8094 |
+
},
|
| 8095 |
+
{
|
| 8096 |
+
"epoch": 0.041025011302809236,
|
| 8097 |
+
"grad_norm": 1.6383607387542725,
|
| 8098 |
+
"learning_rate": 4.097148575848868e-05,
|
| 8099 |
+
"loss": 2.0473,
|
| 8100 |
+
"step": 103700
|
| 8101 |
+
},
|
| 8102 |
+
{
|
| 8103 |
+
"epoch": 0.041304093012352154,
|
| 8104 |
+
"grad_norm": 1.5813179016113281,
|
| 8105 |
+
"learning_rate": 4.095444630473478e-05,
|
| 8106 |
+
"loss": 2.0462,
|
| 8107 |
+
"step": 103800
|
| 8108 |
+
},
|
| 8109 |
+
{
|
| 8110 |
+
"epoch": 0.04158317472189508,
|
| 8111 |
+
"grad_norm": 1.5996896028518677,
|
| 8112 |
+
"learning_rate": 4.093739433756665e-05,
|
| 8113 |
+
"loss": 2.0445,
|
| 8114 |
+
"step": 103900
|
| 8115 |
+
},
|
| 8116 |
+
{
|
| 8117 |
+
"epoch": 0.041862256431438,
|
| 8118 |
+
"grad_norm": 1.627163290977478,
|
| 8119 |
+
"learning_rate": 4.09203298703585e-05,
|
| 8120 |
+
"loss": 2.0512,
|
| 8121 |
+
"step": 104000
|
| 8122 |
+
},
|
| 8123 |
+
{
|
| 8124 |
+
"epoch": 0.041862256431438,
|
| 8125 |
+
"eval_loss": 2.2715542316436768,
|
| 8126 |
+
"eval_runtime": 51.5969,
|
| 8127 |
+
"eval_samples_per_second": 197.57,
|
| 8128 |
+
"eval_steps_per_second": 1.55,
|
| 8129 |
+
"step": 104000
|
| 8130 |
+
},
|
| 8131 |
+
{
|
| 8132 |
+
"epoch": 0.00027908170954291995,
|
| 8133 |
+
"grad_norm": 1.621848225593567,
|
| 8134 |
+
"learning_rate": 4.090325291649436e-05,
|
| 8135 |
+
"loss": 2.056,
|
| 8136 |
+
"step": 104100
|
| 8137 |
+
},
|
| 8138 |
+
{
|
| 8139 |
+
"epoch": 0.0005581634190858399,
|
| 8140 |
+
"grad_norm": 1.5937111377716064,
|
| 8141 |
+
"learning_rate": 4.088616348936804e-05,
|
| 8142 |
+
"loss": 2.0527,
|
| 8143 |
+
"step": 104200
|
| 8144 |
+
},
|
| 8145 |
+
{
|
| 8146 |
+
"epoch": 0.0008372451286287599,
|
| 8147 |
+
"grad_norm": 3.054250478744507,
|
| 8148 |
+
"learning_rate": 4.0869061602383166e-05,
|
| 8149 |
+
"loss": 2.1108,
|
| 8150 |
+
"step": 104300
|
| 8151 |
+
},
|
| 8152 |
+
{
|
| 8153 |
+
"epoch": 0.0011163268381716798,
|
| 8154 |
+
"grad_norm": 1.599531888961792,
|
| 8155 |
+
"learning_rate": 4.0851947268953096e-05,
|
| 8156 |
+
"loss": 2.148,
|
| 8157 |
+
"step": 104400
|
| 8158 |
+
},
|
| 8159 |
+
{
|
| 8160 |
+
"epoch": 0.0013954085477146,
|
| 8161 |
+
"grad_norm": 1.4354475736618042,
|
| 8162 |
+
"learning_rate": 4.083482050250098e-05,
|
| 8163 |
+
"loss": 2.1317,
|
| 8164 |
+
"step": 104500
|
| 8165 |
+
},
|
| 8166 |
+
{
|
| 8167 |
+
"epoch": 0.0016744902572575198,
|
| 8168 |
+
"grad_norm": 1.6972846984863281,
|
| 8169 |
+
"learning_rate": 4.08176813164597e-05,
|
| 8170 |
+
"loss": 2.1313,
|
| 8171 |
+
"step": 104600
|
| 8172 |
+
},
|
| 8173 |
+
{
|
| 8174 |
+
"epoch": 0.00195357196680044,
|
| 8175 |
+
"grad_norm": 1.3760380744934082,
|
| 8176 |
+
"learning_rate": 4.0800529724271896e-05,
|
| 8177 |
+
"loss": 2.1105,
|
| 8178 |
+
"step": 104700
|
| 8179 |
+
},
|
| 8180 |
+
{
|
| 8181 |
+
"epoch": 0.0022326536763433596,
|
| 8182 |
+
"grad_norm": 1.5800480842590332,
|
| 8183 |
+
"learning_rate": 4.0783365739389924e-05,
|
| 8184 |
+
"loss": 2.1108,
|
| 8185 |
+
"step": 104800
|
| 8186 |
+
},
|
| 8187 |
+
{
|
| 8188 |
+
"epoch": 0.0025117353858862797,
|
| 8189 |
+
"grad_norm": 1.6185262203216553,
|
| 8190 |
+
"learning_rate": 4.076618937527585e-05,
|
| 8191 |
+
"loss": 2.0927,
|
| 8192 |
+
"step": 104900
|
| 8193 |
+
},
|
| 8194 |
+
{
|
| 8195 |
+
"epoch": 0.0027908170954292,
|
| 8196 |
+
"grad_norm": 1.666864037513733,
|
| 8197 |
+
"learning_rate": 4.07490006454015e-05,
|
| 8198 |
+
"loss": 2.096,
|
| 8199 |
+
"step": 105000
|
| 8200 |
+
},
|
| 8201 |
+
{
|
| 8202 |
+
"epoch": 0.0027908170954292,
|
| 8203 |
+
"eval_loss": 2.255873441696167,
|
| 8204 |
+
"eval_runtime": 52.3492,
|
| 8205 |
+
"eval_samples_per_second": 194.731,
|
| 8206 |
+
"eval_steps_per_second": 1.528,
|
| 8207 |
+
"step": 105000
|
| 8208 |
+
},
|
| 8209 |
+
{
|
| 8210 |
+
"epoch": 0.00306989880497212,
|
| 8211 |
+
"grad_norm": 1.558935284614563,
|
| 8212 |
+
"learning_rate": 4.0731799563248334e-05,
|
| 8213 |
+
"loss": 2.0891,
|
| 8214 |
+
"step": 105100
|
| 8215 |
+
},
|
| 8216 |
+
{
|
| 8217 |
+
"epoch": 0.0033489805145150396,
|
| 8218 |
+
"grad_norm": 1.62444269657135,
|
| 8219 |
+
"learning_rate": 4.0714586142307546e-05,
|
| 8220 |
+
"loss": 2.0738,
|
| 8221 |
+
"step": 105200
|
| 8222 |
+
},
|
| 8223 |
+
{
|
| 8224 |
+
"epoch": 0.0036280622240579597,
|
| 8225 |
+
"grad_norm": 1.6997108459472656,
|
| 8226 |
+
"learning_rate": 4.069736039607998e-05,
|
| 8227 |
+
"loss": 2.0742,
|
| 8228 |
+
"step": 105300
|
| 8229 |
+
},
|
| 8230 |
+
{
|
| 8231 |
+
"epoch": 0.00390714393360088,
|
| 8232 |
+
"grad_norm": 1.2430105209350586,
|
| 8233 |
+
"learning_rate": 4.0680122338076156e-05,
|
| 8234 |
+
"loss": 2.0892,
|
| 8235 |
+
"step": 105400
|
| 8236 |
+
},
|
| 8237 |
+
{
|
| 8238 |
+
"epoch": 0.0041862256431437995,
|
| 8239 |
+
"grad_norm": 1.6265345811843872,
|
| 8240 |
+
"learning_rate": 4.0662871981816266e-05,
|
| 8241 |
+
"loss": 2.0759,
|
| 8242 |
+
"step": 105500
|
| 8243 |
+
},
|
| 8244 |
+
{
|
| 8245 |
+
"epoch": 0.004465307352686719,
|
| 8246 |
+
"grad_norm": 1.429768681526184,
|
| 8247 |
+
"learning_rate": 4.064560934083012e-05,
|
| 8248 |
+
"loss": 2.0731,
|
| 8249 |
+
"step": 105600
|
| 8250 |
+
},
|
| 8251 |
+
{
|
| 8252 |
+
"epoch": 0.00474438906222964,
|
| 8253 |
+
"grad_norm": 1.5453941822052002,
|
| 8254 |
+
"learning_rate": 4.062833442865719e-05,
|
| 8255 |
+
"loss": 2.0334,
|
| 8256 |
+
"step": 105700
|
| 8257 |
+
},
|
| 8258 |
+
{
|
| 8259 |
+
"epoch": 0.005023470771772559,
|
| 8260 |
+
"grad_norm": 1.6710811853408813,
|
| 8261 |
+
"learning_rate": 4.061104725884654e-05,
|
| 8262 |
+
"loss": 2.0727,
|
| 8263 |
+
"step": 105800
|
| 8264 |
+
},
|
| 8265 |
+
{
|
| 8266 |
+
"epoch": 0.00530255248131548,
|
| 8267 |
+
"grad_norm": 1.668752908706665,
|
| 8268 |
+
"learning_rate": 4.0593747844956896e-05,
|
| 8269 |
+
"loss": 2.0566,
|
| 8270 |
+
"step": 105900
|
| 8271 |
+
},
|
| 8272 |
+
{
|
| 8273 |
+
"epoch": 0.0055816341908584,
|
| 8274 |
+
"grad_norm": 1.6783727407455444,
|
| 8275 |
+
"learning_rate": 4.057643620055654e-05,
|
| 8276 |
+
"loss": 2.0607,
|
| 8277 |
+
"step": 106000
|
| 8278 |
+
},
|
| 8279 |
+
{
|
| 8280 |
+
"epoch": 0.0055816341908584,
|
| 8281 |
+
"eval_loss": 2.251800537109375,
|
| 8282 |
+
"eval_runtime": 51.9592,
|
| 8283 |
+
"eval_samples_per_second": 196.192,
|
| 8284 |
+
"eval_steps_per_second": 1.54,
|
| 8285 |
+
"step": 106000
|
| 8286 |
+
},
|
| 8287 |
+
{
|
| 8288 |
+
"epoch": 0.005860715900401319,
|
| 8289 |
+
"grad_norm": 1.4173957109451294,
|
| 8290 |
+
"learning_rate": 4.055911233922338e-05,
|
| 8291 |
+
"loss": 2.0537,
|
| 8292 |
+
"step": 106100
|
| 8293 |
+
},
|
| 8294 |
+
{
|
| 8295 |
+
"epoch": 0.00613979760994424,
|
| 8296 |
+
"grad_norm": 1.6333705186843872,
|
| 8297 |
+
"learning_rate": 4.054177627454487e-05,
|
| 8298 |
+
"loss": 2.0679,
|
| 8299 |
+
"step": 106200
|
| 8300 |
+
},
|
| 8301 |
+
{
|
| 8302 |
+
"epoch": 0.0064188793194871595,
|
| 8303 |
+
"grad_norm": 1.078372836112976,
|
| 8304 |
+
"learning_rate": 4.0524428020118074e-05,
|
| 8305 |
+
"loss": 2.0599,
|
| 8306 |
+
"step": 106300
|
| 8307 |
+
},
|
| 8308 |
+
{
|
| 8309 |
+
"epoch": 0.006697961029030079,
|
| 8310 |
+
"grad_norm": 1.7360562086105347,
|
| 8311 |
+
"learning_rate": 4.0507067589549595e-05,
|
| 8312 |
+
"loss": 2.0451,
|
| 8313 |
+
"step": 106400
|
| 8314 |
+
},
|
| 8315 |
+
{
|
| 8316 |
+
"epoch": 0.006977042738573,
|
| 8317 |
+
"grad_norm": 1.0331660509109497,
|
| 8318 |
+
"learning_rate": 4.048969499645559e-05,
|
| 8319 |
+
"loss": 2.0569,
|
| 8320 |
+
"step": 106500
|
| 8321 |
+
},
|
| 8322 |
+
{
|
| 8323 |
+
"epoch": 0.0072561244481159195,
|
| 8324 |
+
"grad_norm": 1.7799084186553955,
|
| 8325 |
+
"learning_rate": 4.0472310254461765e-05,
|
| 8326 |
+
"loss": 2.0452,
|
| 8327 |
+
"step": 106600
|
| 8328 |
+
},
|
| 8329 |
+
{
|
| 8330 |
+
"epoch": 0.007535206157658839,
|
| 8331 |
+
"grad_norm": 1.6667675971984863,
|
| 8332 |
+
"learning_rate": 4.045491337720333e-05,
|
| 8333 |
+
"loss": 2.0386,
|
| 8334 |
+
"step": 106700
|
| 8335 |
+
},
|
| 8336 |
+
{
|
| 8337 |
+
"epoch": 0.00781428786720176,
|
| 8338 |
+
"grad_norm": 1.5602868795394897,
|
| 8339 |
+
"learning_rate": 4.043750437832504e-05,
|
| 8340 |
+
"loss": 2.0379,
|
| 8341 |
+
"step": 106800
|
| 8342 |
+
},
|
| 8343 |
+
{
|
| 8344 |
+
"epoch": 0.00809336957674468,
|
| 8345 |
+
"grad_norm": 1.670214295387268,
|
| 8346 |
+
"learning_rate": 4.0420083271481144e-05,
|
| 8347 |
+
"loss": 2.0362,
|
| 8348 |
+
"step": 106900
|
| 8349 |
+
},
|
| 8350 |
+
{
|
| 8351 |
+
"epoch": 0.008372451286287599,
|
| 8352 |
+
"grad_norm": 1.6006022691726685,
|
| 8353 |
+
"learning_rate": 4.040265007033538e-05,
|
| 8354 |
+
"loss": 2.042,
|
| 8355 |
+
"step": 107000
|
| 8356 |
+
},
|
| 8357 |
+
{
|
| 8358 |
+
"epoch": 0.008372451286287599,
|
| 8359 |
+
"eval_loss": 2.252915620803833,
|
| 8360 |
+
"eval_runtime": 51.9529,
|
| 8361 |
+
"eval_samples_per_second": 196.216,
|
| 8362 |
+
"eval_steps_per_second": 1.54,
|
| 8363 |
+
"step": 107000
|
| 8364 |
+
},
|
| 8365 |
+
{
|
| 8366 |
+
"epoch": 0.008651532995830519,
|
| 8367 |
+
"grad_norm": 1.5068638324737549,
|
| 8368 |
+
"learning_rate": 4.0385204788561e-05,
|
| 8369 |
+
"loss": 2.0379,
|
| 8370 |
+
"step": 107100
|
| 8371 |
+
},
|
| 8372 |
+
{
|
| 8373 |
+
"epoch": 0.008930614705373438,
|
| 8374 |
+
"grad_norm": 1.7357958555221558,
|
| 8375 |
+
"learning_rate": 4.0367747439840694e-05,
|
| 8376 |
+
"loss": 2.0289,
|
| 8377 |
+
"step": 107200
|
| 8378 |
+
},
|
| 8379 |
+
{
|
| 8380 |
+
"epoch": 0.00920969641491636,
|
| 8381 |
+
"grad_norm": 1.6686527729034424,
|
| 8382 |
+
"learning_rate": 4.0350278037866654e-05,
|
| 8383 |
+
"loss": 2.0413,
|
| 8384 |
+
"step": 107300
|
| 8385 |
+
},
|
| 8386 |
+
{
|
| 8387 |
+
"epoch": 0.00948877812445928,
|
| 8388 |
+
"grad_norm": 1.226962924003601,
|
| 8389 |
+
"learning_rate": 4.0332796596340485e-05,
|
| 8390 |
+
"loss": 2.0204,
|
| 8391 |
+
"step": 107400
|
| 8392 |
+
},
|
| 8393 |
+
{
|
| 8394 |
+
"epoch": 0.0097678598340022,
|
| 8395 |
+
"grad_norm": 1.696763515472412,
|
| 8396 |
+
"learning_rate": 4.031530312897327e-05,
|
| 8397 |
+
"loss": 2.0296,
|
| 8398 |
+
"step": 107500
|
| 8399 |
+
},
|
| 8400 |
+
{
|
| 8401 |
+
"epoch": 0.010046941543545119,
|
| 8402 |
+
"grad_norm": 1.6187881231307983,
|
| 8403 |
+
"learning_rate": 4.0297797649485515e-05,
|
| 8404 |
+
"loss": 2.0211,
|
| 8405 |
+
"step": 107600
|
| 8406 |
+
},
|
| 8407 |
+
{
|
| 8408 |
+
"epoch": 0.010326023253088039,
|
| 8409 |
+
"grad_norm": 1.082560658454895,
|
| 8410 |
+
"learning_rate": 4.028028017160712e-05,
|
| 8411 |
+
"loss": 2.0304,
|
| 8412 |
+
"step": 107700
|
| 8413 |
+
},
|
| 8414 |
+
{
|
| 8415 |
+
"epoch": 0.01060510496263096,
|
| 8416 |
+
"grad_norm": 1.551513910293579,
|
| 8417 |
+
"learning_rate": 4.026275070907744e-05,
|
| 8418 |
+
"loss": 2.0332,
|
| 8419 |
+
"step": 107800
|
| 8420 |
+
},
|
| 8421 |
+
{
|
| 8422 |
+
"epoch": 0.01088418667217388,
|
| 8423 |
+
"grad_norm": 1.658700704574585,
|
| 8424 |
+
"learning_rate": 4.024520927564521e-05,
|
| 8425 |
+
"loss": 2.0134,
|
| 8426 |
+
"step": 107900
|
| 8427 |
+
},
|
| 8428 |
+
{
|
| 8429 |
+
"epoch": 0.0111632683817168,
|
| 8430 |
+
"grad_norm": 1.734298586845398,
|
| 8431 |
+
"learning_rate": 4.022765588506854e-05,
|
| 8432 |
+
"loss": 2.0259,
|
| 8433 |
+
"step": 108000
|
| 8434 |
+
},
|
| 8435 |
+
{
|
| 8436 |
+
"epoch": 0.0111632683817168,
|
| 8437 |
+
"eval_loss": 2.248534917831421,
|
| 8438 |
+
"eval_runtime": 51.918,
|
| 8439 |
+
"eval_samples_per_second": 196.348,
|
| 8440 |
+
"eval_steps_per_second": 1.541,
|
| 8441 |
+
"step": 108000
|
| 8442 |
+
},
|
| 8443 |
+
{
|
| 8444 |
+
"epoch": 0.011442350091259719,
|
| 8445 |
+
"grad_norm": 1.3358790874481201,
|
| 8446 |
+
"learning_rate": 4.021009055111493e-05,
|
| 8447 |
+
"loss": 2.0226,
|
| 8448 |
+
"step": 108100
|
| 8449 |
+
},
|
| 8450 |
+
{
|
| 8451 |
+
"epoch": 0.011721431800802639,
|
| 8452 |
+
"grad_norm": 1.6433982849121094,
|
| 8453 |
+
"learning_rate": 4.019251328756125e-05,
|
| 8454 |
+
"loss": 2.0231,
|
| 8455 |
+
"step": 108200
|
| 8456 |
+
},
|
| 8457 |
+
{
|
| 8458 |
+
"epoch": 0.012000513510345558,
|
| 8459 |
+
"grad_norm": 1.6571450233459473,
|
| 8460 |
+
"learning_rate": 4.0174924108193734e-05,
|
| 8461 |
+
"loss": 2.0272,
|
| 8462 |
+
"step": 108300
|
| 8463 |
+
},
|
| 8464 |
+
{
|
| 8465 |
+
"epoch": 0.01227959521988848,
|
| 8466 |
+
"grad_norm": 1.5485866069793701,
|
| 8467 |
+
"learning_rate": 4.015732302680795e-05,
|
| 8468 |
+
"loss": 2.0154,
|
| 8469 |
+
"step": 108400
|
| 8470 |
+
},
|
| 8471 |
+
{
|
| 8472 |
+
"epoch": 0.0125586769294314,
|
| 8473 |
+
"grad_norm": 1.717331886291504,
|
| 8474 |
+
"learning_rate": 4.0139710057208794e-05,
|
| 8475 |
+
"loss": 2.0186,
|
| 8476 |
+
"step": 108500
|
| 8477 |
+
},
|
| 8478 |
+
{
|
| 8479 |
+
"epoch": 0.012837758638974319,
|
| 8480 |
+
"grad_norm": 1.3007159233093262,
|
| 8481 |
+
"learning_rate": 4.012208521321049e-05,
|
| 8482 |
+
"loss": 2.017,
|
| 8483 |
+
"step": 108600
|
| 8484 |
+
},
|
| 8485 |
+
{
|
| 8486 |
+
"epoch": 0.013116840348517239,
|
| 8487 |
+
"grad_norm": 1.6857043504714966,
|
| 8488 |
+
"learning_rate": 4.01044485086366e-05,
|
| 8489 |
+
"loss": 2.0183,
|
| 8490 |
+
"step": 108700
|
| 8491 |
+
},
|
| 8492 |
+
{
|
| 8493 |
+
"epoch": 0.013395922058060158,
|
| 8494 |
+
"grad_norm": 1.5786222219467163,
|
| 8495 |
+
"learning_rate": 4.0086799957319965e-05,
|
| 8496 |
+
"loss": 2.0081,
|
| 8497 |
+
"step": 108800
|
| 8498 |
+
},
|
| 8499 |
+
{
|
| 8500 |
+
"epoch": 0.013675003767603078,
|
| 8501 |
+
"grad_norm": 1.708694338798523,
|
| 8502 |
+
"learning_rate": 4.0069139573102715e-05,
|
| 8503 |
+
"loss": 2.0052,
|
| 8504 |
+
"step": 108900
|
| 8505 |
+
},
|
| 8506 |
+
{
|
| 8507 |
+
"epoch": 0.013954085477146,
|
| 8508 |
+
"grad_norm": 1.6401610374450684,
|
| 8509 |
+
"learning_rate": 4.005146736983627e-05,
|
| 8510 |
+
"loss": 1.998,
|
| 8511 |
+
"step": 109000
|
| 8512 |
+
},
|
| 8513 |
+
{
|
| 8514 |
+
"epoch": 0.013954085477146,
|
| 8515 |
+
"eval_loss": 2.2400035858154297,
|
| 8516 |
+
"eval_runtime": 52.1098,
|
| 8517 |
+
"eval_samples_per_second": 195.625,
|
| 8518 |
+
"eval_steps_per_second": 1.535,
|
| 8519 |
+
"step": 109000
|
| 8520 |
+
},
|
| 8521 |
+
{
|
| 8522 |
+
"epoch": 0.01423316718668892,
|
| 8523 |
+
"grad_norm": 1.3219704627990723,
|
| 8524 |
+
"learning_rate": 4.0033783361381324e-05,
|
| 8525 |
+
"loss": 2.0154,
|
| 8526 |
+
"step": 109100
|
| 8527 |
+
},
|
| 8528 |
+
{
|
| 8529 |
+
"epoch": 0.014512248896231839,
|
| 8530 |
+
"grad_norm": 1.7542061805725098,
|
| 8531 |
+
"learning_rate": 4.001608756160781e-05,
|
| 8532 |
+
"loss": 2.0129,
|
| 8533 |
+
"step": 109200
|
| 8534 |
+
},
|
| 8535 |
+
{
|
| 8536 |
+
"epoch": 0.014791330605774759,
|
| 8537 |
+
"grad_norm": 1.690807819366455,
|
| 8538 |
+
"learning_rate": 3.999837998439494e-05,
|
| 8539 |
+
"loss": 2.0031,
|
| 8540 |
+
"step": 109300
|
| 8541 |
+
},
|
| 8542 |
+
{
|
| 8543 |
+
"epoch": 0.015070412315317678,
|
| 8544 |
+
"grad_norm": 1.652106523513794,
|
| 8545 |
+
"learning_rate": 3.9980660643631137e-05,
|
| 8546 |
+
"loss": 2.0025,
|
| 8547 |
+
"step": 109400
|
| 8548 |
+
},
|
| 8549 |
+
{
|
| 8550 |
+
"epoch": 0.015349494024860598,
|
| 8551 |
+
"grad_norm": 1.5204631090164185,
|
| 8552 |
+
"learning_rate": 3.996292955321406e-05,
|
| 8553 |
+
"loss": 2.0024,
|
| 8554 |
+
"step": 109500
|
| 8555 |
+
},
|
| 8556 |
+
{
|
| 8557 |
+
"epoch": 0.01562857573440352,
|
| 8558 |
+
"grad_norm": 1.6795552968978882,
|
| 8559 |
+
"learning_rate": 3.9945186727050574e-05,
|
| 8560 |
+
"loss": 2.0078,
|
| 8561 |
+
"step": 109600
|
| 8562 |
+
},
|
| 8563 |
+
{
|
| 8564 |
+
"epoch": 0.015907657443946437,
|
| 8565 |
+
"grad_norm": 1.7057318687438965,
|
| 8566 |
+
"learning_rate": 3.992743217905678e-05,
|
| 8567 |
+
"loss": 2.0012,
|
| 8568 |
+
"step": 109700
|
| 8569 |
+
},
|
| 8570 |
+
{
|
| 8571 |
+
"epoch": 0.01618673915348936,
|
| 8572 |
+
"grad_norm": 1.5931191444396973,
|
| 8573 |
+
"learning_rate": 3.990966592315793e-05,
|
| 8574 |
+
"loss": 2.0042,
|
| 8575 |
+
"step": 109800
|
| 8576 |
+
},
|
| 8577 |
+
{
|
| 8578 |
+
"epoch": 0.01646582086303228,
|
| 8579 |
+
"grad_norm": 1.6618363857269287,
|
| 8580 |
+
"learning_rate": 3.989188797328851e-05,
|
| 8581 |
+
"loss": 2.0029,
|
| 8582 |
+
"step": 109900
|
| 8583 |
+
},
|
| 8584 |
+
{
|
| 8585 |
+
"epoch": 0.016744902572575198,
|
| 8586 |
+
"grad_norm": 1.6900277137756348,
|
| 8587 |
+
"learning_rate": 3.987409834339211e-05,
|
| 8588 |
+
"loss": 1.9952,
|
| 8589 |
+
"step": 110000
|
| 8590 |
+
},
|
| 8591 |
+
{
|
| 8592 |
+
"epoch": 0.016744902572575198,
|
| 8593 |
+
"eval_loss": 2.2469289302825928,
|
| 8594 |
+
"eval_runtime": 52.1385,
|
| 8595 |
+
"eval_samples_per_second": 195.518,
|
| 8596 |
+
"eval_steps_per_second": 1.534,
|
| 8597 |
+
"step": 110000
|
| 8598 |
+
},
|
| 8599 |
+
{
|
| 8600 |
+
"epoch": 0.01702398428211812,
|
| 8601 |
+
"grad_norm": 1.4866619110107422,
|
| 8602 |
+
"learning_rate": 3.985629704742153e-05,
|
| 8603 |
+
"loss": 2.0031,
|
| 8604 |
+
"step": 110100
|
| 8605 |
+
},
|
| 8606 |
+
{
|
| 8607 |
+
"epoch": 0.017303065991661037,
|
| 8608 |
+
"grad_norm": 1.759400725364685,
|
| 8609 |
+
"learning_rate": 3.9838484099338714e-05,
|
| 8610 |
+
"loss": 1.9986,
|
| 8611 |
+
"step": 110200
|
| 8612 |
+
},
|
| 8613 |
+
{
|
| 8614 |
+
"epoch": 0.01758214770120396,
|
| 8615 |
+
"grad_norm": 1.6036171913146973,
|
| 8616 |
+
"learning_rate": 3.9820659513114735e-05,
|
| 8617 |
+
"loss": 2.0032,
|
| 8618 |
+
"step": 110300
|
| 8619 |
+
},
|
| 8620 |
+
{
|
| 8621 |
+
"epoch": 0.017861229410746877,
|
| 8622 |
+
"grad_norm": 1.650638461112976,
|
| 8623 |
+
"learning_rate": 3.9802823302729806e-05,
|
| 8624 |
+
"loss": 1.9968,
|
| 8625 |
+
"step": 110400
|
| 8626 |
+
},
|
| 8627 |
+
{
|
| 8628 |
+
"epoch": 0.018140311120289798,
|
| 8629 |
+
"grad_norm": 0.8783386945724487,
|
| 8630 |
+
"learning_rate": 3.978497548217324e-05,
|
| 8631 |
+
"loss": 1.927,
|
| 8632 |
+
"step": 110500
|
| 8633 |
+
},
|
| 8634 |
+
{
|
| 8635 |
+
"epoch": 0.01841939282983272,
|
| 8636 |
+
"grad_norm": 1.0009950399398804,
|
| 8637 |
+
"learning_rate": 3.9767116065443464e-05,
|
| 8638 |
+
"loss": 1.8641,
|
| 8639 |
+
"step": 110600
|
| 8640 |
+
},
|
| 8641 |
+
{
|
| 8642 |
+
"epoch": 0.018698474539375638,
|
| 8643 |
+
"grad_norm": 1.0648219585418701,
|
| 8644 |
+
"learning_rate": 3.974924506654801e-05,
|
| 8645 |
+
"loss": 1.8177,
|
| 8646 |
+
"step": 110700
|
| 8647 |
+
},
|
| 8648 |
+
{
|
| 8649 |
+
"epoch": 0.01897755624891856,
|
| 8650 |
+
"grad_norm": 0.8869590759277344,
|
| 8651 |
+
"learning_rate": 3.9731362499503474e-05,
|
| 8652 |
+
"loss": 1.8099,
|
| 8653 |
+
"step": 110800
|
| 8654 |
+
},
|
| 8655 |
+
{
|
| 8656 |
+
"epoch": 0.019256637958461477,
|
| 8657 |
+
"grad_norm": 1.1313343048095703,
|
| 8658 |
+
"learning_rate": 3.971346837833556e-05,
|
| 8659 |
+
"loss": 1.7797,
|
| 8660 |
+
"step": 110900
|
| 8661 |
+
},
|
| 8662 |
+
{
|
| 8663 |
+
"epoch": 0.0195357196680044,
|
| 8664 |
+
"grad_norm": 0.87049400806427,
|
| 8665 |
+
"learning_rate": 3.969556271707898e-05,
|
| 8666 |
+
"loss": 1.7763,
|
| 8667 |
+
"step": 111000
|
| 8668 |
+
},
|
| 8669 |
+
{
|
| 8670 |
+
"epoch": 0.0195357196680044,
|
| 8671 |
+
"eval_loss": 2.235299587249756,
|
| 8672 |
+
"eval_runtime": 52.3725,
|
| 8673 |
+
"eval_samples_per_second": 194.644,
|
| 8674 |
+
"eval_steps_per_second": 1.528,
|
| 8675 |
+
"step": 111000
|
| 8676 |
+
},
|
| 8677 |
+
{
|
| 8678 |
+
"epoch": 0.01981480137754732,
|
| 8679 |
+
"grad_norm": 0.9816115498542786,
|
| 8680 |
+
"learning_rate": 3.967764552977754e-05,
|
| 8681 |
+
"loss": 1.7607,
|
| 8682 |
+
"step": 111100
|
| 8683 |
+
},
|
| 8684 |
+
{
|
| 8685 |
+
"epoch": 0.020093883087090238,
|
| 8686 |
+
"grad_norm": 1.1495822668075562,
|
| 8687 |
+
"learning_rate": 3.9659716830484085e-05,
|
| 8688 |
+
"loss": 1.7387,
|
| 8689 |
+
"step": 111200
|
| 8690 |
+
},
|
| 8691 |
+
{
|
| 8692 |
+
"epoch": 0.02037296479663316,
|
| 8693 |
+
"grad_norm": 1.117358684539795,
|
| 8694 |
+
"learning_rate": 3.9641776633260464e-05,
|
| 8695 |
+
"loss": 1.7432,
|
| 8696 |
+
"step": 111300
|
| 8697 |
+
},
|
| 8698 |
+
{
|
| 8699 |
+
"epoch": 0.020652046506176077,
|
| 8700 |
+
"grad_norm": 1.1056963205337524,
|
| 8701 |
+
"learning_rate": 3.962382495217757e-05,
|
| 8702 |
+
"loss": 1.7173,
|
| 8703 |
+
"step": 111400
|
| 8704 |
+
},
|
| 8705 |
+
{
|
| 8706 |
+
"epoch": 0.020931128215719,
|
| 8707 |
+
"grad_norm": 1.3292206525802612,
|
| 8708 |
+
"learning_rate": 3.960586180131528e-05,
|
| 8709 |
+
"loss": 1.7183,
|
| 8710 |
+
"step": 111500
|
| 8711 |
+
},
|
| 8712 |
+
{
|
| 8713 |
+
"epoch": 0.02121020992526192,
|
| 8714 |
+
"grad_norm": 1.182369589805603,
|
| 8715 |
+
"learning_rate": 3.9587887194762485e-05,
|
| 8716 |
+
"loss": 1.7172,
|
| 8717 |
+
"step": 111600
|
| 8718 |
+
},
|
| 8719 |
+
{
|
| 8720 |
+
"epoch": 0.021489291634804838,
|
| 8721 |
+
"grad_norm": 0.8990152478218079,
|
| 8722 |
+
"learning_rate": 3.956990114661705e-05,
|
| 8723 |
+
"loss": 1.6941,
|
| 8724 |
+
"step": 111700
|
| 8725 |
+
},
|
| 8726 |
+
{
|
| 8727 |
+
"epoch": 0.02176837334434776,
|
| 8728 |
+
"grad_norm": 1.1310220956802368,
|
| 8729 |
+
"learning_rate": 3.955190367098582e-05,
|
| 8730 |
+
"loss": 1.6928,
|
| 8731 |
+
"step": 111800
|
| 8732 |
+
},
|
| 8733 |
+
{
|
| 8734 |
+
"epoch": 0.022047455053890677,
|
| 8735 |
+
"grad_norm": 0.880366325378418,
|
| 8736 |
+
"learning_rate": 3.9533894781984606e-05,
|
| 8737 |
+
"loss": 1.6886,
|
| 8738 |
+
"step": 111900
|
| 8739 |
+
},
|
| 8740 |
+
{
|
| 8741 |
+
"epoch": 0.0223265367634336,
|
| 8742 |
+
"grad_norm": 0.8530257940292358,
|
| 8743 |
+
"learning_rate": 3.951587449373816e-05,
|
| 8744 |
+
"loss": 1.6671,
|
| 8745 |
+
"step": 112000
|
| 8746 |
+
},
|
| 8747 |
+
{
|
| 8748 |
+
"epoch": 0.0223265367634336,
|
| 8749 |
+
"eval_loss": 2.2474164962768555,
|
| 8750 |
+
"eval_runtime": 52.3853,
|
| 8751 |
+
"eval_samples_per_second": 194.597,
|
| 8752 |
+
"eval_steps_per_second": 1.527,
|
| 8753 |
+
"step": 112000
|
| 8754 |
+
},
|
| 8755 |
+
{
|
| 8756 |
+
"epoch": 0.022605618472976517,
|
| 8757 |
+
"grad_norm": 1.0223313570022583,
|
| 8758 |
+
"learning_rate": 3.949784282038018e-05,
|
| 8759 |
+
"loss": 1.6847,
|
| 8760 |
+
"step": 112100
|
| 8761 |
+
},
|
| 8762 |
+
{
|
| 8763 |
+
"epoch": 0.022884700182519438,
|
| 8764 |
+
"grad_norm": 0.878471314907074,
|
| 8765 |
+
"learning_rate": 3.9479799776053306e-05,
|
| 8766 |
+
"loss": 1.6647,
|
| 8767 |
+
"step": 112200
|
| 8768 |
+
},
|
| 8769 |
+
{
|
| 8770 |
+
"epoch": 0.02316378189206236,
|
| 8771 |
+
"grad_norm": 1.2385903596878052,
|
| 8772 |
+
"learning_rate": 3.9461745374909066e-05,
|
| 8773 |
+
"loss": 1.6462,
|
| 8774 |
+
"step": 112300
|
| 8775 |
+
},
|
| 8776 |
+
{
|
| 8777 |
+
"epoch": 0.023442863601605277,
|
| 8778 |
+
"grad_norm": 1.071331262588501,
|
| 8779 |
+
"learning_rate": 3.9443679631107924e-05,
|
| 8780 |
+
"loss": 1.641,
|
| 8781 |
+
"step": 112400
|
| 8782 |
+
},
|
| 8783 |
+
{
|
| 8784 |
+
"epoch": 0.0237219453111482,
|
| 8785 |
+
"grad_norm": 0.8975522518157959,
|
| 8786 |
+
"learning_rate": 3.942560255881922e-05,
|
| 8787 |
+
"loss": 1.6262,
|
| 8788 |
+
"step": 112500
|
| 8789 |
+
},
|
| 8790 |
+
{
|
| 8791 |
+
"epoch": 0.024001027020691117,
|
| 8792 |
+
"grad_norm": 0.8427426218986511,
|
| 8793 |
+
"learning_rate": 3.94075141722212e-05,
|
| 8794 |
+
"loss": 1.6467,
|
| 8795 |
+
"step": 112600
|
| 8796 |
+
},
|
| 8797 |
+
{
|
| 8798 |
+
"epoch": 0.024280108730234038,
|
| 8799 |
+
"grad_norm": 0.8209208250045776,
|
| 8800 |
+
"learning_rate": 3.938941448550098e-05,
|
| 8801 |
+
"loss": 1.6257,
|
| 8802 |
+
"step": 112700
|
| 8803 |
+
},
|
| 8804 |
+
{
|
| 8805 |
+
"epoch": 0.02455919043977696,
|
| 8806 |
+
"grad_norm": 0.8760822415351868,
|
| 8807 |
+
"learning_rate": 3.937130351285452e-05,
|
| 8808 |
+
"loss": 1.6292,
|
| 8809 |
+
"step": 112800
|
| 8810 |
+
},
|
| 8811 |
+
{
|
| 8812 |
+
"epoch": 0.024838272149319877,
|
| 8813 |
+
"grad_norm": 0.9947307705879211,
|
| 8814 |
+
"learning_rate": 3.935318126848664e-05,
|
| 8815 |
+
"loss": 1.6358,
|
| 8816 |
+
"step": 112900
|
| 8817 |
+
},
|
| 8818 |
+
{
|
| 8819 |
+
"epoch": 0.0251173538588628,
|
| 8820 |
+
"grad_norm": 1.6038973331451416,
|
| 8821 |
+
"learning_rate": 3.933504776661102e-05,
|
| 8822 |
+
"loss": 1.8923,
|
| 8823 |
+
"step": 113000
|
| 8824 |
+
},
|
| 8825 |
+
{
|
| 8826 |
+
"epoch": 0.0251173538588628,
|
| 8827 |
+
"eval_loss": 2.267655849456787,
|
| 8828 |
+
"eval_runtime": 51.8322,
|
| 8829 |
+
"eval_samples_per_second": 196.673,
|
| 8830 |
+
"eval_steps_per_second": 1.543,
|
| 8831 |
+
"step": 113000
|
| 8832 |
+
},
|
| 8833 |
+
{
|
| 8834 |
+
"epoch": 0.025396435568405717,
|
| 8835 |
+
"grad_norm": 1.4726253747940063,
|
| 8836 |
+
"learning_rate": 3.931690302145014e-05,
|
| 8837 |
+
"loss": 1.9579,
|
| 8838 |
+
"step": 113100
|
| 8839 |
+
},
|
| 8840 |
+
{
|
| 8841 |
+
"epoch": 0.025675517277948638,
|
| 8842 |
+
"grad_norm": 1.4382061958312988,
|
| 8843 |
+
"learning_rate": 3.9298747047235327e-05,
|
| 8844 |
+
"loss": 1.9466,
|
| 8845 |
+
"step": 113200
|
| 8846 |
+
},
|
| 8847 |
+
{
|
| 8848 |
+
"epoch": 0.025954598987491556,
|
| 8849 |
+
"grad_norm": 1.5668888092041016,
|
| 8850 |
+
"learning_rate": 3.928057985820668e-05,
|
| 8851 |
+
"loss": 1.9479,
|
| 8852 |
+
"step": 113300
|
| 8853 |
+
},
|
| 8854 |
+
{
|
| 8855 |
+
"epoch": 0.026233680697034478,
|
| 8856 |
+
"grad_norm": 1.5239572525024414,
|
| 8857 |
+
"learning_rate": 3.926240146861314e-05,
|
| 8858 |
+
"loss": 1.9306,
|
| 8859 |
+
"step": 113400
|
| 8860 |
+
},
|
| 8861 |
+
{
|
| 8862 |
+
"epoch": 0.0265127624065774,
|
| 8863 |
+
"grad_norm": 1.6437016725540161,
|
| 8864 |
+
"learning_rate": 3.924421189271239e-05,
|
| 8865 |
+
"loss": 1.9364,
|
| 8866 |
+
"step": 113500
|
| 8867 |
+
},
|
| 8868 |
+
{
|
| 8869 |
+
"epoch": 0.026791844116120317,
|
| 8870 |
+
"grad_norm": 1.632081151008606,
|
| 8871 |
+
"learning_rate": 3.9226011144770904e-05,
|
| 8872 |
+
"loss": 1.9297,
|
| 8873 |
+
"step": 113600
|
| 8874 |
+
},
|
| 8875 |
+
{
|
| 8876 |
+
"epoch": 0.02707092582566324,
|
| 8877 |
+
"grad_norm": 1.577589750289917,
|
| 8878 |
+
"learning_rate": 3.920779923906393e-05,
|
| 8879 |
+
"loss": 1.9223,
|
| 8880 |
+
"step": 113700
|
| 8881 |
+
},
|
| 8882 |
+
{
|
| 8883 |
+
"epoch": 0.027350007535206156,
|
| 8884 |
+
"grad_norm": 1.5894604921340942,
|
| 8885 |
+
"learning_rate": 3.918957618987545e-05,
|
| 8886 |
+
"loss": 1.9263,
|
| 8887 |
+
"step": 113800
|
| 8888 |
+
},
|
| 8889 |
+
{
|
| 8890 |
+
"epoch": 0.027629089244749078,
|
| 8891 |
+
"grad_norm": 1.5186114311218262,
|
| 8892 |
+
"learning_rate": 3.9171342011498185e-05,
|
| 8893 |
+
"loss": 1.9261,
|
| 8894 |
+
"step": 113900
|
| 8895 |
+
},
|
| 8896 |
+
{
|
| 8897 |
+
"epoch": 0.027908170954292,
|
| 8898 |
+
"grad_norm": 1.639272689819336,
|
| 8899 |
+
"learning_rate": 3.9153096718233604e-05,
|
| 8900 |
+
"loss": 1.9166,
|
| 8901 |
+
"step": 114000
|
| 8902 |
+
},
|
| 8903 |
+
{
|
| 8904 |
+
"epoch": 0.027908170954292,
|
| 8905 |
+
"eval_loss": 2.2479372024536133,
|
| 8906 |
+
"eval_runtime": 51.8659,
|
| 8907 |
+
"eval_samples_per_second": 196.545,
|
| 8908 |
+
"eval_steps_per_second": 1.542,
|
| 8909 |
+
"step": 114000
|
| 8910 |
+
},
|
| 8911 |
+
{
|
| 8912 |
+
"epoch": 0.028187252663834917,
|
| 8913 |
+
"grad_norm": 1.5154080390930176,
|
| 8914 |
+
"learning_rate": 3.913484032439187e-05,
|
| 8915 |
+
"loss": 1.916,
|
| 8916 |
+
"step": 114100
|
| 8917 |
+
},
|
| 8918 |
+
{
|
| 8919 |
+
"epoch": 0.02846633437337784,
|
| 8920 |
+
"grad_norm": 1.5147068500518799,
|
| 8921 |
+
"learning_rate": 3.911657284429189e-05,
|
| 8922 |
+
"loss": 1.9153,
|
| 8923 |
+
"step": 114200
|
| 8924 |
+
},
|
| 8925 |
+
{
|
| 8926 |
+
"epoch": 0.028745416082920756,
|
| 8927 |
+
"grad_norm": 1.5977636575698853,
|
| 8928 |
+
"learning_rate": 3.9098294292261205e-05,
|
| 8929 |
+
"loss": 1.9027,
|
| 8930 |
+
"step": 114300
|
| 8931 |
+
},
|
| 8932 |
+
{
|
| 8933 |
+
"epoch": 0.029024497792463678,
|
| 8934 |
+
"grad_norm": 1.4557409286499023,
|
| 8935 |
+
"learning_rate": 3.908000468263609e-05,
|
| 8936 |
+
"loss": 1.9073,
|
| 8937 |
+
"step": 114400
|
| 8938 |
+
},
|
| 8939 |
+
{
|
| 8940 |
+
"epoch": 0.0293035795020066,
|
| 8941 |
+
"grad_norm": 1.5657786130905151,
|
| 8942 |
+
"learning_rate": 3.9061704029761495e-05,
|
| 8943 |
+
"loss": 1.9026,
|
| 8944 |
+
"step": 114500
|
| 8945 |
+
},
|
| 8946 |
+
{
|
| 8947 |
+
"epoch": 0.029582661211549517,
|
| 8948 |
+
"grad_norm": 1.5239813327789307,
|
| 8949 |
+
"learning_rate": 3.904339234799098e-05,
|
| 8950 |
+
"loss": 1.9092,
|
| 8951 |
+
"step": 114600
|
| 8952 |
+
},
|
| 8953 |
+
{
|
| 8954 |
+
"epoch": 0.02986174292109244,
|
| 8955 |
+
"grad_norm": 1.4764059782028198,
|
| 8956 |
+
"learning_rate": 3.9025069651686816e-05,
|
| 8957 |
+
"loss": 1.9019,
|
| 8958 |
+
"step": 114700
|
| 8959 |
+
},
|
| 8960 |
+
{
|
| 8961 |
+
"epoch": 0.030140824630635357,
|
| 8962 |
+
"grad_norm": 1.5580713748931885,
|
| 8963 |
+
"learning_rate": 3.9006735955219874e-05,
|
| 8964 |
+
"loss": 1.8958,
|
| 8965 |
+
"step": 114800
|
| 8966 |
+
},
|
| 8967 |
+
{
|
| 8968 |
+
"epoch": 0.030419906340178278,
|
| 8969 |
+
"grad_norm": 1.5598928928375244,
|
| 8970 |
+
"learning_rate": 3.898839127296968e-05,
|
| 8971 |
+
"loss": 1.8961,
|
| 8972 |
+
"step": 114900
|
| 8973 |
+
},
|
| 8974 |
+
{
|
| 8975 |
+
"epoch": 0.030698988049721196,
|
| 8976 |
+
"grad_norm": 1.4871125221252441,
|
| 8977 |
+
"learning_rate": 3.897003561932434e-05,
|
| 8978 |
+
"loss": 1.8978,
|
| 8979 |
+
"step": 115000
|
| 8980 |
+
},
|
| 8981 |
+
{
|
| 8982 |
+
"epoch": 0.030698988049721196,
|
| 8983 |
+
"eval_loss": 2.2493135929107666,
|
| 8984 |
+
"eval_runtime": 51.934,
|
| 8985 |
+
"eval_samples_per_second": 196.288,
|
| 8986 |
+
"eval_steps_per_second": 1.54,
|
| 8987 |
+
"step": 115000
|
| 8988 |
+
},
|
| 8989 |
+
{
|
| 8990 |
+
"epoch": 0.030978069759264117,
|
| 8991 |
+
"grad_norm": 1.4677025079727173,
|
| 8992 |
+
"learning_rate": 3.89516690086806e-05,
|
| 8993 |
+
"loss": 1.8905,
|
| 8994 |
+
"step": 115100
|
| 8995 |
+
},
|
| 8996 |
+
{
|
| 8997 |
+
"epoch": 0.03125715146880704,
|
| 8998 |
+
"grad_norm": 1.4988230466842651,
|
| 8999 |
+
"learning_rate": 3.8933291455443786e-05,
|
| 9000 |
+
"loss": 1.8935,
|
| 9001 |
+
"step": 115200
|
| 9002 |
+
},
|
| 9003 |
+
{
|
| 9004 |
+
"epoch": 0.03153623317834996,
|
| 9005 |
+
"grad_norm": 1.5289825201034546,
|
| 9006 |
+
"learning_rate": 3.891490297402781e-05,
|
| 9007 |
+
"loss": 1.8875,
|
| 9008 |
+
"step": 115300
|
| 9009 |
+
},
|
| 9010 |
+
{
|
| 9011 |
+
"epoch": 0.031815314887892875,
|
| 9012 |
+
"grad_norm": 1.480161428451538,
|
| 9013 |
+
"learning_rate": 3.889650357885514e-05,
|
| 9014 |
+
"loss": 1.8897,
|
| 9015 |
+
"step": 115400
|
| 9016 |
+
},
|
| 9017 |
+
{
|
| 9018 |
+
"epoch": 0.0320943965974358,
|
| 9019 |
+
"grad_norm": 1.6153870820999146,
|
| 9020 |
+
"learning_rate": 3.887809328435683e-05,
|
| 9021 |
+
"loss": 1.8878,
|
| 9022 |
+
"step": 115500
|
| 9023 |
+
},
|
| 9024 |
+
{
|
| 9025 |
+
"epoch": 0.03237347830697872,
|
| 9026 |
+
"grad_norm": 1.409119725227356,
|
| 9027 |
+
"learning_rate": 3.8859672104972454e-05,
|
| 9028 |
+
"loss": 1.8976,
|
| 9029 |
+
"step": 115600
|
| 9030 |
+
},
|
| 9031 |
+
{
|
| 9032 |
+
"epoch": 0.032652560016521635,
|
| 9033 |
+
"grad_norm": 1.5816664695739746,
|
| 9034 |
+
"learning_rate": 3.884124005515015e-05,
|
| 9035 |
+
"loss": 1.8876,
|
| 9036 |
+
"step": 115700
|
| 9037 |
+
},
|
| 9038 |
+
{
|
| 9039 |
+
"epoch": 0.03293164172606456,
|
| 9040 |
+
"grad_norm": 1.5077482461929321,
|
| 9041 |
+
"learning_rate": 3.882279714934657e-05,
|
| 9042 |
+
"loss": 1.888,
|
| 9043 |
+
"step": 115800
|
| 9044 |
+
},
|
| 9045 |
+
{
|
| 9046 |
+
"epoch": 0.03321072343560748,
|
| 9047 |
+
"grad_norm": 1.4968522787094116,
|
| 9048 |
+
"learning_rate": 3.880434340202686e-05,
|
| 9049 |
+
"loss": 1.8841,
|
| 9050 |
+
"step": 115900
|
| 9051 |
+
},
|
| 9052 |
+
{
|
| 9053 |
+
"epoch": 0.033489805145150396,
|
| 9054 |
+
"grad_norm": 1.467680811882019,
|
| 9055 |
+
"learning_rate": 3.878587882766472e-05,
|
| 9056 |
+
"loss": 1.8832,
|
| 9057 |
+
"step": 116000
|
| 9058 |
+
},
|
| 9059 |
+
{
|
| 9060 |
+
"epoch": 0.033489805145150396,
|
| 9061 |
+
"eval_loss": 2.2356553077697754,
|
| 9062 |
+
"eval_runtime": 51.966,
|
| 9063 |
+
"eval_samples_per_second": 196.167,
|
| 9064 |
+
"eval_steps_per_second": 1.539,
|
| 9065 |
+
"step": 116000
|
| 9066 |
+
},
|
| 9067 |
+
{
|
| 9068 |
+
"epoch": 0.033768886854693314,
|
| 9069 |
+
"grad_norm": 1.4993896484375,
|
| 9070 |
+
"learning_rate": 3.87674034407423e-05,
|
| 9071 |
+
"loss": 1.8783,
|
| 9072 |
+
"step": 116100
|
| 9073 |
+
},
|
| 9074 |
+
{
|
| 9075 |
+
"epoch": 0.03404796856423624,
|
| 9076 |
+
"grad_norm": 1.479200005531311,
|
| 9077 |
+
"learning_rate": 3.8748917255750225e-05,
|
| 9078 |
+
"loss": 1.8864,
|
| 9079 |
+
"step": 116200
|
| 9080 |
+
},
|
| 9081 |
+
{
|
| 9082 |
+
"epoch": 0.03432705027377916,
|
| 9083 |
+
"grad_norm": 1.4365732669830322,
|
| 9084 |
+
"learning_rate": 3.873042028718764e-05,
|
| 9085 |
+
"loss": 1.877,
|
| 9086 |
+
"step": 116300
|
| 9087 |
+
},
|
| 9088 |
+
{
|
| 9089 |
+
"epoch": 0.034606131983322075,
|
| 9090 |
+
"grad_norm": 1.4553158283233643,
|
| 9091 |
+
"learning_rate": 3.871191254956208e-05,
|
| 9092 |
+
"loss": 1.8873,
|
| 9093 |
+
"step": 116400
|
| 9094 |
+
},
|
| 9095 |
+
{
|
| 9096 |
+
"epoch": 0.034885213692865,
|
| 9097 |
+
"grad_norm": 1.505732774734497,
|
| 9098 |
+
"learning_rate": 3.8693394057389574e-05,
|
| 9099 |
+
"loss": 1.8737,
|
| 9100 |
+
"step": 116500
|
| 9101 |
+
},
|
| 9102 |
+
{
|
| 9103 |
+
"epoch": 0.03516429540240792,
|
| 9104 |
+
"grad_norm": 1.4430396556854248,
|
| 9105 |
+
"learning_rate": 3.8674864825194574e-05,
|
| 9106 |
+
"loss": 1.8743,
|
| 9107 |
+
"step": 116600
|
| 9108 |
+
},
|
| 9109 |
+
{
|
| 9110 |
+
"epoch": 0.035443377111950836,
|
| 9111 |
+
"grad_norm": 1.5058810710906982,
|
| 9112 |
+
"learning_rate": 3.865632486750996e-05,
|
| 9113 |
+
"loss": 1.8734,
|
| 9114 |
+
"step": 116700
|
| 9115 |
+
},
|
| 9116 |
+
{
|
| 9117 |
+
"epoch": 0.035722458821493754,
|
| 9118 |
+
"grad_norm": 1.551901936531067,
|
| 9119 |
+
"learning_rate": 3.8637774198877e-05,
|
| 9120 |
+
"loss": 1.8754,
|
| 9121 |
+
"step": 116800
|
| 9122 |
+
},
|
| 9123 |
+
{
|
| 9124 |
+
"epoch": 0.03600154053103668,
|
| 9125 |
+
"grad_norm": 1.5766918659210205,
|
| 9126 |
+
"learning_rate": 3.86192128338454e-05,
|
| 9127 |
+
"loss": 1.878,
|
| 9128 |
+
"step": 116900
|
| 9129 |
+
},
|
| 9130 |
+
{
|
| 9131 |
+
"epoch": 0.036280622240579596,
|
| 9132 |
+
"grad_norm": 1.5308256149291992,
|
| 9133 |
+
"learning_rate": 3.860064078697323e-05,
|
| 9134 |
+
"loss": 1.8833,
|
| 9135 |
+
"step": 117000
|
| 9136 |
+
},
|
| 9137 |
+
{
|
| 9138 |
+
"epoch": 0.036280622240579596,
|
| 9139 |
+
"eval_loss": 2.2324624061584473,
|
| 9140 |
+
"eval_runtime": 51.883,
|
| 9141 |
+
"eval_samples_per_second": 196.48,
|
| 9142 |
+
"eval_steps_per_second": 1.542,
|
| 9143 |
+
"step": 117000
|
| 9144 |
}
|
| 9145 |
],
|
| 9146 |
"logging_steps": 100,
|
|
|
|
| 9160 |
"attributes": {}
|
| 9161 |
}
|
| 9162 |
},
|
| 9163 |
+
"total_flos": 1.0210871976394752e+19,
|
| 9164 |
"train_batch_size": 128,
|
| 9165 |
"trial_name": null,
|
| 9166 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5777
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da5f5517b1675eb630da2afe2ee47f40a6f105aba3407f1e48d33a873836c026
|
| 3 |
size 5777
|